Skip to main content

Full text of "CMH Pub 5-7-1 Campaign in the Marianas"

See other formats

: I ««W-*- - 


Philip A. Crowl 


The War in the Pacific 


Philip A. Crowl 



WASHINGTON, llC, 1 99 ? 

Library of Congress Catalog Card Number: 60-60000 

lirst Primed I960— CMH Pub 5-7-1 

Stetson Conn, General Editor 

Advisory Committee 

(As of 15 March 1959) 

Elmer Ellis Maj. Gen. Hugh M. Harris 

University of Missouri U.S. Continental Army Command 

Samuel Flagg Bemis grig. Gen. Edgar C. Dokrnan 

Yale University Array War College 

Gordon A. Craig Brig. Gen. Frederick R. Zierath 

Princeion University Command and General Staff College 

Oron J. Hale B r j g Gen. Kenneth F. Zitzman 

University of Virginia Industrial College of the Armed Forces 

W. Stull Holt Col Vincent J. Esposito 

University of Washington United States Military Academy 

T. Harry Williams 
Louisiana State University 

Office of the Chief of Military History 

Col. Warren H. Hoover, Acting Chief 

„..,.,. Stetson Conn 

Chief Historian , , . 

_,. , . ,*. — . . . Lt. Col. Joseph Kockis 

Chief, Histories Division , j_ ' ' ». i 

_,.-_,,.._... Lt. Col. E.. L, btcck 

Chict, Publication Division x , 

_,. _,. . Joseph R. Friedman 

Editor in Chief p ir n 

Chief, Cartographic Branch 

Chief, Photographic Branch argare ac ey 

to Those Who Served 


In the capture of the southern Marianas, including the recapture of Guam, 
during the summer of 1944, Army ground and air forces played an important, 
though subordinate, role to that of the Navy and its Marine Corps. Marine 
personnel constituted the bulk of the combat troops employed. The objective of 
this campaign was "to secure control of sea communications through the 
Central Pacific by isolating and neutralizing the Carolines and by the establish- 
ment of sea and air bases for operations against Japanese sea routes and 
long-range air attacks against the Japanese home land." Its success would pro- 
vide steppingstones from which the Americans could threaten further attack 
westward toward the Philippines, Formosa, and Japan itself, and would gain 
bases from which the Army Air Forces' new very long range bombers, the 
B-29's, could strike at Japan's heartland. Recognizing and accepting the chal- 
lenge, the Japanese Navy suffered heavy and irreplaceable losses in the 
accompanying Battle of the Philippine Sea ; and the islands after capture became 
the base for all the massive air attacks on Japan, beginning in November 1944. 

In the operations described in the present volume, landings against strong 
opposition demonstrated the soundness of the amphibious doctrine and tech- 
niques evolved out of hard experience in preceding Pacific operations. Bitter 
inland fighting followed the landings, with Army and Marine Corps divisions 
engaged side by side. The author's account and corresponding Marine Corps 
histories of these operations provide ample opportunity to study the differences 
in the fighting techniques of the two services. Dr. Crowl also deals frankly with 
one of the best-known controversies of World War II, that of Smith versus 
Smith, but concludes that it was the exception to generally excellent interservice 

With team effort among the military services the order of the day, this record 
of the Army's experience in working with the Navy and the Marine Corps 
should be particularly valuable both now and in the future. 

Washington, D.C. Colonel, U.S.A. 

16 March 1959 Acting Chief of Military History 


The Author 

Philip A. Crowl, who has an M.A. from the State University of Iowa and a 
Ph.D. from the Johns Hopkins University, taught History at the Johns Hopkins 
University and at Princeton. Commissioned in the Navy in World War II, he 
became a lieutenant (senior grade) and commanding officer of an LCI gunboat 
that was in action at Leyte Gulf, Lingayen Gulf, and Okinawa. From 1949 
through 1956, Dr. Crowl was a historian with the Office, Chief of Military 
History. He was awarded the James V. Forrestal Fellowship for 1953-54 t0 
study command relationships in amphibious warfare in World War II. Since 
1957 he has been in Department of State Intelligence. Dr: Crowl is author of 
Maryland During and After the Revolution (1943) and coauthor of The U.S. 
Marines and Amphibious War ( 1951 ) and Seizure of the Gilberts and Marshalls 



This volume is a companion piece to Seizure of the Gilberts and Marshalls 
by Philip A. Crowl and Edmund G, Love, also published in the Pacific subseries 
of the UNITED STATES ARMY IN WORLD WAR II. Together, the two 
volumes cover the beginning and climax (although not the conclusion) of the 
Central Pacific phase of the war against Japan, with special emphasis, of 
course, on the U.S. Army's contribution to the victories won in that area. 1 
Specifically, Campaign in the Marianas treats of the capture of Saipan, Tinian, 
and Guam in the southern Marianas; the strategic and tactical plans leading 
thereto; supporting operations by naval and air units; and the final development 
and exploitation of these islands as bases for furtherance of American joint 
operations against the Japanese homeland. 

The word joint cannot be overemphasized in connection with any considera- 
tion of U.S. operations in the Central Pacific. It was predominantly a U.S. 
Navy theater under the command of Admiral Chester W. Nimitz. The main 
burden of the amphibious and ground fighting in the Marianas, as in the 
Gilberts and Marshalls, fell on the shoulders of the U.S. Marine Corps, whose 
troops far outnumbered those of the U.S. Army. The author recognizes this and 
recognizes also that, by concentrating on the activities of the Army, this volume 
in a sense presents a distorted picture. The distortion is deliberate. The book 
represents, by definition, one segment of the history of the U.S. Army in World 
War II. Excellent official and semiofficial histories of U.S. Navy and U.S. 
Marine Corps operations in the Marianas have already been published. The 
present narrative of Army activities should add in some measure to what has 
already been written about the campaign. The reader may also gain additional 
insight into the nature of joint operations and interscrvice co-ordination. 

Because the number of Army troops participating in the Marianas Campaign 
was comparatively small, it has been possible to devote more attention here to 
small unit actions than in the volumes of the series that deal with the move- 
ments of great armies and corps over large continental land masses. In much 
of this narrative the spotlight centers on the infantry company. Ideally, as 
much attention should have been devoted to equivalent artillery units, especially 
since Army artillery played a major role in the Marianas Campaign. Un- 
fortunately, the records kept by artillery units during the campaign were — to 
understate the matter — terse. Unfortunately also, Army field historians who 

1 The conclusion of this phase of the war is covered in Roy E. Appleman, James M, 
Burns, Russell A. Gugcler, and John T. Stevens, Okinawa: The Last Battle, UNITED 
STATES ARMY IN WORLD WAR II (Washington, 1948). 


accompanied the troops and who were to supplement and correct the official 
records by conducting on-the-spot interviews, mostly neglected the artillery in 
favor of the more mobile infantry. 

No really adequate acknowledgment can be made to the many kind and 
industrious people who helped to bring this book to completion. On substantive 
matters of strategy and tactics, Dr, Louis Morton, Chief of the Pacific Section, 
Office of the Chief of Military History, during the preparation of this volume, 
was a tireless and able critic. On questions of literary style, Dr. Kent Roberts 
Greenfield, formerly Chief Historian, Department of the Army, was the same. 
The final editing of the manuscript was performed by Miss Mary Ann Bacon, 
whose eye for detail is remarkable. None of them is responsible for any errors 
of fact and interpretation or gaucheries of style I may have persisted in despite 
their stern warnings. To all I am very grateful. 

I owe a special debt of gratitude to Mr. Wsevolod Aglaimoff whose services 
went far above and beyond the call of his duties as Deputy Chief Historian for 
Cartography. His maps speak for themselves. What is not so apparent is his 
meticulous scrutiny of the written narrative presented here and his expert ad- 
vice on all matters tactical — advice based on almost a half-century's intensive 
study of military tactics. 

Maj. Gen. A. C. Smith, formerly Chief of Military History, Col. George G. 
O'Connor, formerly Chief of War Histories Division, and the military members 
of their staffs were liberal in their technical assistance and made it possible for 
me to have access to the classified source material upon which this volume is 
based. Mr. Thomas Wilds is responsible for unearthing, in some cases translating, 
and in most cases interpreting the widely scattered and often obscure Japanese 
source material upon which is based the account of enemy plans, defensive 
preparations, and combat activity. Miss Margaret Plumb began and Mr. Stanley 
L. Falk completed a highly useful study of the Guam phase of the campaign. 
Mr. Falk's excellent draft narrative of the operations of the 77th Infantry 
Division on Guam forms the basis for the account of that division's actions 
presented here. Loretto Carroll Stevens was copy editor for the volume and 
Norma Heacock Sherris was responsible for the selection of photographs. The 
Index was compiled by Nicholas J. Anthony. Mr. Israel Wice, Chief Archivist, 
General Reference Branch, and his staff and Miss Lois Aldrich of the World 
War II Records Division, NARS, were unfailingly patient and courteous. The 
debt to officers of the Army, Navy, and Marine Corps who read and criticized 
various portions of the manuscript or corresponded with the author is acknowl- 
edged in the Bibliographical Note appended to the volume. 

Since so much of this volume deals with Marine Corps operations, I have 
been particularly dependent on the co-operation of members of the Historical 
Branch, G-3 Division, Headquarters, U.S. Marine Corps. Special thanks are 
due to the late Lt. Col. Frank O. Hough, USMCR, Lt. Col, Harry Edwards, 
USMC, Lt. Col. Carl W. Hoffman, USMC, and Maj. O. R. Lodge, USMC. 

Washington, D.C. PHILIP A. GROWL 

16 March 1959 




Chapter Page 


Prewar Origins of the Central Pacific Concept ...... 2 

War in the Pacific: First Year 4 

Revival of the Central Pacific Concept 6 

General MacArthur's Strategy 9 

Enter the Army Air Forces 1 1 

Cairo Conference 12 

Scheduling Operations . 13 

Acceleration of Operations 14 

Washington Planning Conferences: February~March IQ44 . 13 


In History 21 

The Japanese Invasion of Guam 22 

Description of the Islands 24 




Organization and Composition of the Attack Force .... 33 

Tactical Planning 39 

Training and Rehearsals 45 

Loading and Embarkation 47 

The Prospects Ahead: Intelligence of the Enemy 50 


Chapter Page 


Prewar Japanese Activities in the Marianas 53 

From Pearl Harbor to Invasion 55 

Japanese Doctrine for Island Defense 62 

Enemy Troop Strength and Dispositions on Saipan .... 64 

Japanese Expectations 69 


Softening the Target: Pre-D-Day Bombardment ..... 7 1 

D-Day Bombardment and Ship-to-Shore Movement .... 78 

Breakdown of the Landing Plan . 85 

Expanding the Beachhead 87 

Summary of the Situation at Nightfall 93 


Counterattack; Night of 15-16 June 95 

Consolidating the Beachhead: 16 June .,,,,,.. 96 

Night of 16-17 June 98 

Change of Plans 98 

First Landings of the 27th Infantry Division 99 

D Plus 3: 17 June 101 

D Plus 3: 18 June 1 1 1 

The Japanese Situation 116 


Battle of the Philippine Sea - 119 

Logistics 123 

Postlanding Naval Gunfire Support 129 

Ctose Air Support 13 l 

Artillery 133 


Action of ig June 138 

Action of 20 June 14 1 

Action of 31 June 144 

Change of Plan: Relief of the 165th Infantry on Nafutan 

Point 148 

Action of 33 June 150 

Stalemate on Nafutan: 33-24 June 151 

Nafutan Secured: 25-38 June 155 



Preparations for the Drive to the North 

22 June; The Jump-off 

23 June: Into Death Valley . . . 
The First Night in Death Valley . 
25 June: Marines on the Flanks . . 

24 June; Action of the 2Jth Division 
24 June: Action on the Flanks . . 




Relief of Major General Ralph C. Smith 191 

Interservice Controversy 192 

Conclusions 197 


25 June 303 

The Plight of the Japanese an 

26 June 313 

2J June 216 

Japanese Reactions 221 

28 June 221 

2g June 227 

30 June 230 

Central Saipan; Sum-up 232 


Drive to Tanapag 235 

Change of Direction . 244 

5 My 244 

6 July 347 

7 My 256 

Final Victory 362 




Plan for the Invasion 271 

The Enemy 278 


Chapter Page 


Preliminary Bombardment 285 

The Landings 288 

Japanese Counterattack: 34—2$ July 293 

Capture of Northern Tinian 294 

Drive to the South 296 

Tinian Secured 301 




The Island , 

Plans for the Invasion . 

Change of Plans 

yyth Infantry Division Training and Preparation 

Loading and Embarkation 

Preliminary Bombardment 

Intelligence of the Enemy 



Troops and Troop Dispositions . 

Supporting Weapons 


Japanese Situation on the Eve of Battle 



W-Day Preliminary Bombardment 339 

From Ship to Shore 342 

The Northern Beaches 343 

The Southern Beaches , 344 

Landing the 305th Infantry 345 

Japanese Counterattack . 347 

Consolidating the Southern Beachhead: 22—24 J u ty • 34^ 

Landing the Reserves 354 

Consolidating the Northern Beachhead: 23—24 J u h .... 356 

Initial Supply Over the Beaches 357 


Chapter Page 


Preparations for the Assault on Orote: 25 July 362 

The Fight in the North; 25 July 363 

Japanese Counterattack: 35—26 July ......... 364 

The Capture of Orote 367 

The Capture of Fonte and the Force Beachhead Line ... 371 

Reconnaissance of Southern Guam 374 


The Japanese Withdrawal . 377 

Drive to the 0~2 Line; 31 July-i August 378 

Supply Problems 383 

To Barrigada and the O—3 Line: 2-4 August 386 

yyth Division: 3 August 398 

Tjth Division: 4 August ,,,,.. 402 

The Marines: 2-4 August 407 


JJth Division: 5-6 August . . . 409 

3d Marine Division: 5—6 August 417 

Capture of Mount Santa Rosa: 7—8 August , 417 

The Marines: j—8 August 434 

The End on Guam 436 





A. Tactical Units of Northern Troops and Landing Force on Saipan . . 449 

B. Pacific Ocean Area Unit of Fire for Ground Weapons 452 

C. Japanese Order of Battle on Saipan 453 

D. Troop List of the 77th Infantry Division for the Guam Operation . . 455 

E. Order of Battle of III Amphibious Corps (less 77th Infantry 

Division) for the Guam Operation 456 





INDEX 475 


i. Estimated Strength of the Japanese Garrison on Tinian 
2. Artillery on Guam 




1. Task Organization for Major Commands for Attack on Saipan 

and Tinian 

2. Task Organization for Major Commands for Attack on Guam . 



1. Pacific Ocean Areas, 15 June 1944 

2. Southern Marianas 

3. Japanese Defense Sectors on Saipan 
4- D Day on Saipan, 15 June 1944 . 

5. Advance 16—18 June 

6. Advance 19-22 June ..... 

7. Into Death Valley, 23-24 June 1944 

8. 27th Division, 25 June 1944 . 

9. 27th Division, 26-27 June 1944 . 
10- 27th Division, 28-30 June 1944 . 
n. Drive to Tanapag, 1-4 July 1944 . 

12. 105th Infantry, Morning, 6 July 1944 

13. 105th Infantry, Afternoon, 6 July 1944 

14. Last Days of Battle on Saipan, 7-9 July 1944 

15. Japanese Defense Sectors on Tinian . . . 

16. Capture of Tinian, 24 July-i August 1944 

17. Disposition of Japanese Troops on Guam, July 1944 

18. Landings on Guam, 21 July 1944 . , , . 

19. Capture of Orote Peninsula, 25-29 July 1944 

20. Securing the Force Beachhead Line, 26—29 J u ty '944 

21. Reconnaissance of Southern Guam, 28 July-2 August 

22. Drive to O-2 Line, 31 July-i August 1944 . . . 
23- Approach to Barrigada, 2 August 1944 

24. Advance to O-3 Line, 2-4 August 1944 .... 

25. Advance to O-4 Line, 5-6 August 1944 .... 

26. Advance 7-8 August and Capture of Mt. Santa Rosa 













2 49 
2 55 
3 6 7 




Maps I-VII Are in Accompanying Map Envelope 

I. Saipan Island 

II. 27th Division, 16-22 June 1944 

III. Advance 5-6 July 

IV. Tinian Island 
V. Island of Guam 

VI. Northern Beachhead, 2 1-25 July 1944 
VII. Southern Beachhead, 2 1-25 July 1944 



Top Naval Commanders in the Marianas Campaign 36 

Vice Adm. Richmond Kelly Turner 37 

Lt. Gen. Holland M. Smith 3 8 

Maj. Gen. Ralph C. Smith 39 

Destruction at Charan Kanoa 75 

Dummy Searchlight Platform on East Coast of Saipan 77 

Japanese Type 96 25-mm, Machine Cannon 78 

Amphibian Tractors in Line Abreast Formation 81 

Congestion on 2d Marine Division Beach 84 

Early Wave on the Beach - 88 

Marines Digging Foxholes 90 

D-Day Command Post 91 

Narrow-Gauge Railroad Near Charan Kanoa 102 

Soldiers Watch Destruction of a Pillbox 103 

Reinforcements Moving Inland 104 

105th Infantrymen Wading in From the Reef 108 

Aslito Field Becomes Conroy Field 113 

Examining an Enemy Gun 115 

Shore Parties Unloading Supplies on Blue Beach 1 125 

Caterpillar Tractor Pulling Ammunition Pallets From an LCM .... 126 

Pontoon Causeway and Barge in Charan Kanoa Harbor 127 

North Central Saipan 134 

Seaward Cliff Line, Nafutan Peninsula . 138 

27th Division Troops at Cliff Edge, Nafutan Peninsula 140 

Infantryman at the Base of Cliff, Nafutan Peninsula 141 

Infantryman Milking an Island Goat 145 

Skirmish Line 156 

Americans Rescuing Baby From Nafutan Cave 158 

Hill 500 164 

Marines on the Crest of Hill 500 . 165 



Japanese Type 93 13,2-rnm. Machine Gun 166 

Japanese Field of Fire 167 

Mt. Tapotchau Dominating Death Valley 174 

Tree Line in Death Valley 175 

Driven to Concealment . 176 

Marines Emerging From Purple Heart Ridge Complex 180 

Tank-Infantry Co-operation 185 

In Hell's Pocket Area 189 

Inland of Garapan Harbor 190 

Maj. Gen. Sanderford Jarman 193 

Lt, Gen. Holland M. Smith 194 

Lt. Gen. Robert C. Richardson, Jr 195 

Maj. Gen. Ralph C. Smith 197 

Lower End of Death Valley 308 

Bazooka Team Preparing to Fire 214 

Maj. Gen. George W. Griner 225 

Truck-Mounted Rocket Launchers Firing 229 

Marines Dash Through Blazing Garapan 239 

Infantrymen Observe Hill 767 240 

Troops Searching Out the Enemy 243 

Harakiri Gulch 246 

Waiting To Move Up 252 

Marines Examining Destroyed Enemy Tank 260 

Flame Thrower Blasting Paradise Valley Cave 263 

Coastal Area, Northwest Tinian 270 

Marianas Leaders Confer at Tinian 273 

LVT With Ramp 276 

155-mrn. Gun Firing at Tinian From Saipan 287 

Invasion Craft 292 

4th Marines Wading Toward White Beach 1 293 

Ushi Point Airfield 296 

Open Terrain of Central Tinian 297 

155-mm, Howitzer Emplacement on Tinian . , 298 

Cliff Line at Tip of Orote Peninsula 308 

Orote Peninsula 309 

Maj. Gen. Roy S. Geiger 311 

Japanese Beach Obstacles at Guam 336 

Japanese Open-Trench Beach Defenses at Agat 337 

First Wave of Landing Craft Heads for Agat Beaches 340 

Asan's Green and Blue Beaches 341 

3d Marine Division Beachhead 343 

Circling Landing Craft 346 

4th Marines Moves Inland Toward Mt. Alifan 350 

Assembly Area of 305th Infantry on 22 July 351 


Smouldering Japanese Tanks . 353 

Maj. Gen. Andrew D. Bruce 354 

Pontoon Barge With Crane 358 

Troops in Bivouac 362 

Japanese Airfield, the Prime Objective on Orote Peninsula 366 

Marine Corps Officers 370 

Antitank Crew 373 

Men of Company B, 305th RCT 381 

Forward Observers of the 77th Division 383 

Artillery Column Moving Inland 385 

Mt. Santa Rosa 418 

Burning U.S. Medium Tanks 426 

Lt. Gen. Alexander A. Vandegrift 435 

LST's, LCI's, Small Boats 442 

Harmon Field, Guam 444 

Photographs are from Department of Defense files with the following excep- 
tions: photographs on pages 37 and 38 from the personal files of Col. Robert C 
Richardson, III; the photograph on page 418 from the Bishop Museum, Oahu, 
Hawaii; and that on page 444, by J. R. Eyerman-Life, (c) 1950 Time Inc. 

The U.S. Army Center of Military History 

The Center of Military History prepares and publishes histories as re- 
quired by the U.S. Army. It coordinates Army historical matters, including 
historical properties, and supervises the Army museum system. It. also 
maintains liaison with public and private agencies and individuals to stimu- 
late interest and study in the field of military history. The Center is located 
at 1099 14th Street, N.W., Washington, D.C. 20005-3402. 



Background of Strategy 

On 15 June 1944 American forces in- 
vaded the island of Saipan thus piercing 
the first hole in the inner line of island 
fortifications that the Japanese had labori- 
ously constructed in order to defend their 
homeland, their empire, and their recent 
conquests in the western Pacific and in 
Asia. Saipan is 1,270 nautical miles from 
Tokyo, 1,430 from Manila, 1,640 from 
Shanghai, and 3,350 from Honolulu. Lo- 
cated in the southern portion of the Mari- 
anas chain, it was the most heavily fortified 
of that group of islands and was consid- 
ered by the Japanese to be a keystone in 
the defensive system for the homeland. 

After twenty-four days of strenuous bat- 
tle and much bloodletting on the part of 
both victor and vanquished, Saipan was 
conquered. On 1 August the little island 
of Tinian, just a few miles to the south, 
fell to U.S. forces, and in the same month 
Guam, the southernmost of the Marianas 
chain, was recaptured from the Japanese, 
who had wrested it from the Americans 
during the first days of the war. 

Speaking of the fall of Saipan, Fleet Ad- 
miral Osami Nagano, Supreme Naval Ad- 
visor to the Emperor, could only remark, 
"Hell is on us." 1 Premier Hideki Tojo 
publicly announced, "Japan has come to 

face an unprecedentcdly great national 
crisis." The following month Tojo resigned 
in disgrace along with his entire war cabi- 
net. His resignation marked a major turn- 
ing point in the war. Up to that time the 
military clique, led and symbolized by 
Tojo, had been in secure control of the 
machinery of government and had dictated 
Imperial policy without any effective re- 
straints. Thereafter, an opposition party 
with strong inclinations toward terminat- 
ing the war made gradual but steady in- 
roads into the councils of state, until at 
last it was able to persuade the Emperor 
to surrender. The loss of Saipan and the 
overthrow of Tojo gave this peace party its 
first opportunity. 2 

The spark that set off this interesting 
train of events was a directive issued by 
the U.S. Joint Chiefs of Staff on \i March 
1944/ This body — consisting of Admiral 
William D. Leahy, President Franklin D. 
Roosevelt's chief of staff; Admiral Ernest 
J. King, Chief of Naval Operations and 
Commander in Chief, U.S. Fleet; General 
George C Marshall, Chief of Staff, U.S. 
Army; and General Henry H. Arnold, 

1 United States Strategic Bombing Survey 
(USSBS) (Pacific), Naval Analysis Division, In- 
terrogations of Japanese Officials, a vols. (Wash- 
ington, 1946) (hereafter cited as USSBS, Inter- 
rogations), 11, 356. 

-USSBS (Pacific), Naval Analysis Division, 
The Campaigns of the Pacific War (Washington, 
1946) (hereafter cited as USSBS Campaigns), p. 
220; USSBS, Chairman's Office, Japan's Struggle 
to End the War (Washington, 1946) pp. 2-3; 
Rohert J. Butow, Japan's Decision to Surrender 
(Stanford, Calif., Stanford University Press, 
194.5), PP- 2 6- 2 9) el passim. 

:t Msg, JCS to MacArthur and Nimitz, 5171 
and 989, 12 Mar 44, CM-OUT 5137. 


Commanding General, U.S. Army Air 
Forces — was responsible, under the Presi- 
dent and in conjunction with its British 
counterpart, for the strategic direction of 
World War II. On 12 March it ordered 
Admiral Chester W. Nimitz, Commander 
in Chief, U.S. Pacific Fleet and Pacific 
Ocean Areas, to occupy the southern 
Marianas beginning 15 June next. The ob- 
jective as stated was "to secure control of 
sea communications through the Central 
Pacific by isolating and neutralizing the 
Carolines and by the establishment of sea 
and air bases for operations against Japa- 
nese sea routes and long range air attacks 
against the Japanese home land." 

The 1 2 March directive was the product 
of a slow if not always steady growth, 
emerging only after a long and sometimes 
bitter conflict of strategic ideas, military 
Interests, and personalities. A leading issue 
of this conflict was what can best be called 
the "Central Pacific concept" of American 
strategy in a war against Japan. 

Prewar Origins of the 
Central Pacific Concept 

Shortly after the termination of World 
War I, when Japan's pretensions in the 
western Pacific and the Far East were be- 
coming steadily more apparent, American 
strategic planners set to work to examine 
possible ways and means of defeating the 
Japanese Empire in the event of war be- 
tween the two nations. From these delib- 
erations emerged a series of plans, dating 
from 1924 through 1938, entitled the 
Orange plans. 4 The product of joint 

Army-Navy effort, these were issued by 
the Joint Army and Navy Board, the pre- 
decessor to the Joint Chiefs of Staff. 

Although the several plans varied in de- 
tail, certain assumptions remained fairly 
constant. The basic concept of the war 
against Japan, as expressed by the Joint 
Board in 1929, was that it would be "pri- 
marily naval in character throughout, un- 
less large Army forces arc employed in 
major land operations in the Western Pa- 
cific, directed toward the isolation and ex- 
haustion of Japan, through control of her 
vital sea communications and through ag- 
gressive operations against her armed 
forces and her economic life." r> To con- 
duct such a war, the U.S. Fleet must first 
move west from Hawaii and establish an 
advanced naval base in the Philippines, 
preferably at Manila Bay if it could be 
held. Such an advance would be either a 
direct movement or a step-by-step process 
involving seizure and occupation of key 
Japanese Mandated Islands in the Mar- 
shall and Carolines, depending upon the 
nature and extent of the enemy resistance. 
From a naval base in the Philippines, it 
was presumed that Japanese trade routes 
through the South China Sea could be cut 
and Japan's economic life throttled. Amer- 
ican forces might also move north to estab- 
lish still more bases in the Ryukyus and 
other islands neighboring Japan, from 
which American naval control could be 
exercised over Japanese home waters and 
American aircraft could harass the home- 
land itself. 

i JU 335, Ser. 22B, 15 Aug 34, sub: Joint Army 
and Navy Basic War Plan, Orange; Ser. sBo, 14 
Jun 38-11 Jan 29; Ser. 546, 9 May 35; Scr. 570, 
19 May 36; Ser. 618, 28 Feb 38. For further dis- 
cussion of prewar strategic plans, see Louis 

Morton, Strategy and Command: Turning the 
Tide, 1941-1943, a forthcoming volume in the 

5 JB 325, Ser. 280, 14 Jun 28-11 Jan 29, sub: 
Joint Army and Navy Basic War Plan, Orange 
— Estimate of the Situation, p, 39, 



The plans assumed it would be essential 
to establish subsidiary U.S. bases in the 
Japanese Mandates, especially in the Mar- 
shall, for the purpose of protecting the 
line of communications between the Phil- 
ippines and the continental United States. 
It was generally agreed that bases in the 
Marshalls, and probably in the Carolines, 
would have to be occupied by U.S. forces 
either in offensive operations in advance of 
a fleet approach to the Philippines or as a 
defensive measure to protect the line of 
communications of American forces oper- 
ating in the western Pacific. The Marianas 
figured only incidentally in the scheme, 
since they lay north of the main route of 
advance from Hawaii to the Philippines. 
Thus first emerged the Central Pacific 
concept of strategy. 

The Orange plans were based on the 
assumption that the United States alone 
would be engaged in a war with Japan. 
With the outbreak of war in Europe in 
1939, and with the gradual strengthening 
of bonds between the United States and 
the anti-Axis nations, especially Great Brit- 
ain, this assumption no longer held. It be- 
came necessary to agree to some combined 
strategic measures in anticipation of the 
day when the United States might active- 
ly enter the war against the Axis. 

As a consequence, British and American 
military staff representatives met in Wash- 
ington in early 1941 to discuss possible 
strategy should the United States become 
a belligerent. 8 In their final report, ABC- 
i, 7 the representatives agreed first that 

"should the United States be compelled to 
resort to war," both nations would con- 
sider the Atlantic and European area to be 
"the decisive theater," since Germany was 
the predominant member of the Axis Pow- 
ers. Thus was enunciated the doctrine of 
"beat Germany first" that prevailed until 
the surrender of Germany on 8 May 1945. 
Until the Germans were defeated, Allied 
strategy in the Far East would be primar- 
ily defensive. American and British forces 
would defend Hong Kong, the Philip- 
pines, and the Netherlands Indies and hold 
Malaya, Singapore, and Java against Jap- 
anese attack. Within this defensive pattern, 
the U.S. Navy was assigned the specific 
offensive mission of capturing positions in 
the Marshalls and preparing to establish 
control over the entire Caroline and Mar- 
shall Islands area. 8 Thus was restated the 
main principle of the Orange plans: hold 
the Philippines, if possible, and gain con- 
trol over the islands and waters of the 
Central Pacific west of Hawaii. 

In April 1941 the U.S. Joint Board set 
about bringing its own plans up to date in 
the light of these American-British Con- 
versations. The new strategic plan, entitled 
Rainbow 5, merely restated the decisions 
of ABC- 1 and assigned more specific tasks 
to the U.S. forces. Germany was to be 
beaten first; the Philippines were to be 
held as long as possible; the U.S. Fleet was 
to prepare to capture positions in the Mar- 
shalls and the Carolines. 8 Thus, once 

6 ABC-i, 27 Mar 41, in Pearl Harbor Attack: 
Hearings before the Joint Committee on the In- 
vestigation of the Pearl Harbor Attack (Washing- 
ion, 1946) (hereafter cited as Pearl Harbor Hear- 
ings), Part 15, pp. 1 48 7- 1 550. 

1 ABC stands for American-British Conversa- 

8 Pearl Harbor Hearings, Part 15, pp. 1511-12, 
1516, 1518-19. 

» Ibid., Part 18, pp. 2885^ 2889-90, 2894, 2919, 
2916. Rainbow 5 was approved by the Secre- 
taries of War and Navy, and although never for- 
mally approved by the President, it was in fart 
the war plan that went into effect on 7 December 


again, the role of the Central Pacific in 
the forthcoming war against Japan was 

War in the Pacific; First Year 

The rapid succession of Japanese vic- 
tories after Japan's attack on Pearl Harbor 
made it impossible for the United States 
and her allies to put into immediate exe- 
cution any of the prewar plans. Guam fell 
to the invaders, followed by Hong Kong, 
Singapore, Manila, and the Philippine Is- 
lands. In short order the United States 
and Great Britain were stripped of all the 
usable advanced bases they once possessed. 
The Netherlands Indies, Malaya, and 
Burma were overrun and the Bismarck 
Archipelago-New Guinea-Solomons area 
was invaded. By May of 1 942 the Japanese 
were well ensconced in a far-flung system 
of mutually supporting bases including the 
Kurils, the Marianas, the Marshafls, the 
Carolines, the Palaus, and Rabaul (in 
New Britain), with outposts in the Gil- 
berts, the Solomons, and New Guinea. 

Flushed with victory, the Japanese high 
command decided to ride its good fortune 
to the limit and push on to Port Moresby, 
New Guinea— the very threshold of Aus- 
tralia; to New Caledonia, Fiji, and Samoa, 
astride the sea lanes between Australia 
and the United States; and to Midway 
and the western Aleutians. The first plan 
was frustrated by the Allied naval victory 
in the Coral Sea in early May 1942, while 
the second died aborning. The Japanese 
gained a tenuous foothold in the Aleutians, 
but in attempting to seize Midway they 
suffered their first decisive defeat of the 
war at the hands of the U.S. Pacific Fleet 
in the great naval and air battle of 3-4 

June 1942.'" The time had come for Al- 
lied counteraction. 

The United States, which had assumed 
major responsibility for the war in the Pa- 
cific, had laid the groundwork for future 
offensive operations against the enemy. On 
30 March 1942 the Joint Chiefs of Staff, 
with the approval of President Roosevelt, 
organized the Pacific theater into two 
commands — the Southwest Pacific Area 
(SWPA) and the Pacific Ocean Areas 
(POA). The former fell to the command 
of General Douglas MacArthur, with 
headquarters in Australia, and included 
Australia, New Guinea, the Bismarck Ar- 
chipelago, the Solomons, the Netherlands 
Indies (except Sumatra), the Philippines, 
and adjacent waters. 11 The Pacific Ocean 
Areas was to be commanded by Admiral 
Nimitz, whose headquarters was at Pearl 
Harbor and who was concurrently Com- 
mander in Chief, U.S. Pacific Fleet. This 
theater included virtually the entire re- 
mainder of the Pacific Ocean. Within its 
boundaries lay the Hawaiian Islands, Mid- 
way, Wake, the Gilberts, the Marshalls, 
the Carolines, the Palaus, the Marianas, 
the Bonins, the Ryukyus, Formosa (Tai- 
wan), and the Japanese home islands. 12 
Because of the immensity of the theater, it 
was subdivided into three areas, North, 
Central, and South Pacific. Nimitz di- 
rectly commanded the first two, but as- 

10 See Samuel Eliot Morison, History of United 
States Naval Operations in World War II, Vol. 
IV, Coral Sea, Midway and Submarine Actions, 
May 1 942-August 1942 (Boston, Little, Brown 
and Company, 1949). 

11 Memo, Marshall and King for President, 30 
Mar 42, no sub, with two inclosures, Directive to 
CINCPOA and Directive to the Supreme Com- 
mander, SWPA, ABC 323.31 POA (1-29-42), 

12 Excluded from POA were the waters off the 
west coast of South America. 


u, a, s, r. 


15 June 1944 

— — Subdivision eoufDARiES 

1000 sooo 


MAP 1 

*? ■ j'o- hntfQ^gi 

signed the third to a subordinate com- 
mander who from October 1942 to June 
1944 was Admiral William F. Halsey, Jr. is 
The first task facing MacArthur and 
Nimitz after the Japanese fleet had been 
turned back at Midway was to render more 
secure the line of communications between 

13 Fleet Admiral William F. Hatsey, USN, and 
Lieutenant Commander J. Bryan, III, USNR, 
Admiral Halsey's Story (New York and London, 
Whittlesey House, McGraw-Hill Book Company, 

Inc., 1947)- 

the United States and Australia. On 2 
July 1942 the Joint Chiefs of Staff ordered 
their Pacific commanders to advance 
through the Solomons and New Guinea 
and seize the Japanese stronghold of Ra- 
baul. On 1 August 1942 the area under 
MacArthur's control was reduced slightly 
by moving his eastward boundary line 
from i6o°E to i59°E from the equator 
southward. This in effect took the lower 
Solomon Islands including Tulagi, Guad- 
alcanal, Florida, the Russells, Malaita, and 


San Cri stobal out o f MacArthur's juris- 


There followed the 
lengthy and exhausting Guadalcanal and 
Papua Campaigns. By January of 1943 
the line of communications was safe, 
even though Rabaul remained in enemy 
hands. 14 

The Japanese were at last on the de- 
fensive; the losses at Pearl Harbor were 
more than replaced by the naval repair 
and construction program. Allied military 
planners could now direct their attention 
to devising ways and means of taking up 
the offensive, and it is not surprising that 
their thoughts turned once again to the 
prewar plans for the Pacific. 

Revival of the Central Pacific Concept 
The Casablanca Conference 

In January 1943 President Roosevelt 
and the U.S. Joint Chiefs of Staff met 
with Prime Minister Winston S. Churchill 
and the British Chiefs of Staff at Casa- 
blanca, French Morocco. There, the Amer- 
ican and British Chiefs, known collectively 

14 See John Miller, jr., Guadalcanal: The First 
Offensive (Washington, 1949), and Samuel Mil- 
ner, Victory in Papua (Washington, 1957), both 
WAR II series; Samuel Eliot Morison, History of 
United States Naval Operations in World War 
II, Vol. IV, Coral Sea, Midway and Submarine 
Actions, and Vol. V, The Struggle for Guadal- 
canal, August ig^s-February 1943 (Boston, Lit- 
tle, Brown and Company, 1949); Maj. John L. 
Zimmerman, USMCR, The Guadalcanal Cam^ 
paign, Historical Division, Division of Public In- 
formation, Headquarters, U.S. Marine Corps 
(Washington, 1949); Wesley Frank Craven and 
James Lea Cate, eds., The Army Air Forces in 
World War II, Vol. IV, The Pacific: Guadal- 
canal to Saipan, August 194s to July 1944 (here- 
after cited as Craven and Cate, AAF IV) (Chi- 
cago, The University of Chicago Press, 1950). 

as the Combined Chiefs of Staff, mapped 
out as best they could the main lines of 
global strategy for the coming year. 15 In 
the midst of prolonged and sometimes 
acrimonious discussion concerning forth- 
coming operations in Europe and the 
Mediterranean, proposed offensives in Bur- 
ma, and aid to China, Admiral King 
stepped forth to assume the role that he 
was to fill for the next two years — that of 
the leading advocate of a greater effort in 
the Pacific and more specifically of a Cen- 
tral Pacific thrust against the Japanese. 

King addressed himself primarily to the 
problem of where to go after Rabaul was 
captured— an operation that was optimis- 
tically assumed to be already well ad- 
vanced with the campaigns in Guadal- 
canal and Papua drawing to a successful 
close. He adhered to the concept of the 
Orange plans and urged that the first 
main Allied objective in the Pacific be the 
Philippines, since they lay athwart the 
line of communications between Japan 
and the oil-rich East Indies and since their 
occupation by Allied forces would permit 
that line to be cut. The best route to this 
objective, he claimed, lay through the 
Central Pacific. Such a drive would in- 
volve "establishing a base in the north- 
western Marshalls and then proceeding to 
Truk and the Marianas." "The Marianas" 
he added, "are the key of the situation be- 

ls For excellent discussions of the Casablanca 
Conference, see Lt. Grace P. Hayes, USN, The 
War Against Japan, Part I, Ch. X, MS, JCS 
Hist Sec; John Miller, jr., "The Casablanca Con- 
ference and Pacific Strategy," Military Affairs, 
XIII (Winter, 1949); Fleet Admiral Ernest J. 
King and Walter M, Whitehill, Fleet Admiral 
King, A Naval Record (New York, W. W. Nor- 
ton & Company, 1952), Ch. XXXIII; Maurice 
Matloff, Strategic Planning for Coalition War- 
fare: 1943-1944, UNITED STATES ARMY IN 
WORLD WAR II (Washington, 1959). 


cause of their location on the Japanese 
line of communication." 16 

An important addition was thus intro- 
duced into the Orange concept of the 
war in the Pacific. The Marianas, which 
had received little attention from the Joint 
Board in its prewar plans, now emerged 
as a major objective in the mind of Ad- 
miral King. The other participants at the 
Casablanca meeting were too concerned 
about more immediate problems to pay 
much attention to King's remarks about 
the Marianas, but the conferees did en- 
dorse planning for a drive through the 
Central Pacific in 1943. 

On 17 January the U.S. Joint Staff 
Planners, a subcommittee of the Joint 
Chiefs of Staff, presented a fourfold pro- 
gram for the forthcoming year's operations 
in the Pacific : ( 1 ) seizure of the Solo- 
mons, of eastern New Guinea as far as Lae 
and Salamaua, and of the New Britain- 
New Ireland area; (2) seizure of Kiska 
and Agattu in the Aleutians 5(3) after the 
fall of Rabaul, seizure and occupation of 
the Gilberts, the Marshalls, and the Caro- 
lines through Truk, and extension of the 
occupation of New Guinea to the border 
of Netherlands New Guinea; and (4) op- 
erations in Burma designed to keep China 
in the war and increase the employment 
of China-based aircraft against Japanese 
shipping. 17 

Most of this program was acceptable to 
the Combined Chiefs, although they stipu- 
lated that the advance in the Central Pa- 
cific should not be allowed to prejudice 
the recapture of Burma. The final agree- 
ment at Casablanca authorized plans to 

be made for a campaign in the Aleutians, 
an advance northwest from Samoa along 
the Solomons-New Britain-New Ireland- 
New Guinea axis to protect the line 
of communications between the United 
States and Australia, diversionary attacks 
against the Malay Barrier, and a Central 
Pacific advance west against the Truk- 
Guam line. 18 

TRIDENT Conference 

In May of 1943 the President and 
the Prime Minister with their Combined 
Chiefs of Staff convened again, this time 
in Washington at the Trident Confer- 
ence. 19 Most of the discussion again cen- 
tered around questions concerning the 
Mediterranean theater and Burma and 
China, but during the course of the meet- 
ing, the Combined Chiefs reaffirmed their 
determination to get on with the war in 
the Central Pacific. 20 

On 14 May the American representa- 
tives circulated to the Combined Chiefs a 
paper drawn up by the various subcom- 
mittees of the Joint Chiefs, entitled Strate- 
gic Plan for the Defeat of Japan. 21 This 
paper was more of an estimate than a plan, 

16 Min, 56th mtg CCS, 14 Jan 43. 

17 CCS 153 (Rev.), 17 Jan 43, title: Situation 
To Be Created in Eastern Theater (Pacific and 
Burma) in 1943. 

18 CCS 170/3, 23 Jan 43, title: Final Rpt to 
President and Prime Minister Summarizing Deci- 
sions by CCS in Casablanca Conf. 

1B For a more detailed discussion of the Tri- 
dent conference, see Philip A. Crowl and Edmund 
G. Love, Seizure of the Gilberts and Marshalls, 
II (Washington, 1955), pp. 12-17; Matloff, Stra- 
tegic Planning for Coalition Warfare, 1g4.3-1g4.jt; 
Hayes, The War Against Japan, Sec. IV, Part II, 
Ch. XIII. 

30 See Charles F. Romanus and Riley Sunder- 
land, Stilwell's Mission to China, UNITED 
ington, 1953), PP. 32 7-33- 

21 CCS 220, 14 May 43; Min, 86th mtg JCS, 
20 May 43; JSSC 40/2, 3 Apr 43; JPS 67/4, 28 



but it did articulate more clearly than be- 
fore the main strategic principles endorsed 
by the highest U.S. planners for the war 
in the Pacific in 1943-44. It was assumed 
that to bring about the unconditional sur- 
render of Japan it would first be necessary 
for the Allies to secure a foothold in China 
in order to make best use of the enormous 
Chinese manpower and to provide air 
bases from which to bomb the Japanese 
mainland, China could be entered by three 
routes: through Burma, through the Strait 
of Malacca and the South China Sea to 
Hong Kong from the west, and across the 
Pacific through the Celebes Sea to Hong 
Kong from the cast. The British, with 
American and Chinese aid, should be held 
responsible for operations along the first 
two routes. The United States would as- 
sume major responsibility for the third. 

Next came the question of how Ameri- 
can forces could best get to the Celebes 
Sea and Hong Kong from positions held 
by the United States. The American plan- 
ners proposed a two-pronged drive by U.S. 
forces, one westward from Hawaii through 
the Central Pacific, the other west and 
north along the Solomons-Bismarck-New 
Guinea line in General MacArthur's 
Southwest Pacific Area. The two thrusts 
were to merge in the Philippines-South 
China Sea area and join in the descent 
upon Hong Kong. In determining priori- 
tics as between the two drives, it was de- 
clared that the main effort in the west- 

Apr 43; JCS 287/1, 8 May 43. All these bear the 
title Strategic Plan for the Defeat of Japan or a 
similar one The subordinate committees of the 
JCS included the Joint War Plans Committee 
(JWPC), the Joint Strategic Survey Committee 
(JSSC), and the Joint Staff Planners (JPS). For 
more detailed discussion of the paper Strategic 
Plan for the Defeat of Japan, see Growl and Love, 
Gilberts and Marshalh, pp. 12-17. 

ward advance should be made through 
the Central Pacific and a subsidiary role 
assigned to the South and Southwest 

Once again Admiral King took the floor 
to press the argument in favor of the Cen- 
tral Pacific drive and more particularly to 
champion his favorite project, the invasion 
of the Marianas. For years, he said, officers 
at the Naval War College in Newport had 
been studying the problem of supporting 
or recovering the Philippines as the sine 
qua non of defeating Japan. Their con- 
clusions all pointed to the route straight 
through the Pacific from the Hawaiian Is- 
lands as the best approach. The Marianas, 
he insisted, were the key to the wes- 
tern Pacific. A major offensive there, he 
claimed, would seriously jeopardize Japa- 
nese lines of communications, most prob- 
ably force the Japanese Fleet into a decisive 
naval engagement, and provide bases from 
which to bomb the Japanese home is- 
lands. 22 

In its final session at Trident the Com- 
bined Chiefs of Staff, although not com- 
mitting themselves on the question of the 
Marianas, agreed to American recommen- 
dations for a two-pronged attack across 
the Pacific. Specifically, they listed the 
following strategic objectives in the Pacific 
for 1943-44: 

1. Conducting air operations in and 
from China, 

2. Ejection of the Japanese from the 

3. Seizure of the Marshall and the Car- 
oline Islands, 

4. Seizure of the Solomons, the Bis- 
marck Archipelago, and Japanese-held 
New Guinea, and 

22 Min, gad mtg CCS, 21 May 43; King and 
Whitehill, Fleet Admiral King, pp. 437~3 8 j 444- 


5. Intensification of operations against 
enemy lines of communication. 

Two months later the Joint Chiefs of 
Staff decided to modify the plans for 
launching the Central Pacific drive. A di- 
rect assault on the Marshalls from Hawaii, 
they reasoned, would require more ship- 
ping and troops than were immediately 
available in the Central Pacific and might 
necessitate draining General MacArthur's 
Southwest Pacific Area of some of its re- 
sources. Also, it would be well to have 
better aerial photographs of the Marshalls 
before attempting the dangerous amphibi- 
ous assault on these Japanese strongholds 
about which almost nothing was known. 
Air bases from which photographic mis- 
sions could be flown should therefore be 
built close to the Marshalls. Largely for 
these reasons the Joint Chiefs, on 20 July, 
ordered Nimitz to capture Tarawa, in the 
Gilberts, and Nauru as a preliminary to 
going into the Marshalls themselves. The 
target date selected was 15 November 

I943- 23 

In August the Combined Chiefs, meet- 
ing at Quebec for the Quadrant Confer- 
ence, accepted the revision as well as a 
schedule of operations proposed by the 
American representatives for the period 
after the capture of the Marshalls. For the 
Central Pacific this included the capture 
of Ponape, Woleai, and Truk in the Caro- 
lines, the development of Truk as a major 
fleet base, and the seizure of Yap and the 
Palaus. At Admiral King's suggestion, an 
invasion of the Marianas was included in 
this program as a possible alternative to 

the Palaus or as a concurrent operation. 24 
A few days after this decision was made 
Admiral Nimitz suggested one further 
change in the plan for initiating the Cen- 
tral Pacific drive. He proposed substi- 
tuting Makin in the Gilberts for Nauru 
because Makin could more easily be as- 
saulted and because it was closer to the 
Marshalls. The proposal was accepted, 
and Nimitz was authorized to seize Tara- 
wa, Makin, and Apamama in the Gil- 
berts. 25 On 20 November 1943 simultan- 
eous amphibious landings were launched 
against Makin and Tarawa by elements 
of the 27th Infantry Division and the 2d 
Marine Division, respectively. Within four 
days both atolls were captured, following 
which Apamama was occupied. 28 The 
Central Pacific drive was under way. 

General MacArthur's Strategy 

Meanwhile, halfway around the world 
at his headquarters in Brisbane, General 
MacArthur was developing strategic plans 
that were not always consonant with the 
ideas prevailing among high echelon plan- 

ts Min, 97th mtg JCS, ao Jul 43; JCS 386/2, 
ao Jul 43, title: Strategy in the Pacific; Rad, JCS 
to CINCPAC, 20 Jul 43, CM-IN 14465. See also 
Crowl and Love, Gilberts and Marshalls, pp. 18- 


24 JCS 446, 5 Aug 43, title: Specific Opns in 
the Pacific and Far East, 1943-44; Min, roist 
mtg JCS, 7 Aug 43; CCS 3 OI /3> 27 Au S 43, 
title: Specific Opns in the Pacific and Far East; 
CQS 319/5, 24 Aug 43, title: Final Rpt to Presi- 
dent and Prime Minister. 

25 Msg, Nimitz to JCS, 260439, 26 Sep 43; 
Memo, Roberts for Handy, 37 Sep 43, Sub: Sub- 
stitution of Makin for Nauru in Operation Gal- 
vanic, ABC 384 Pacific (28 Jun 43); Memo, 
U.S. CofS for COMINCH, 27 Sep 43, same sub; 
OPD Exec a, Item ib. 

2B See Crowl and Love, Gilberts and Marshalls; 
Historical Division, War Department, The Cap- 
ture of Makin, AMERICAN FORCES IN AC- 
TION (Washington, 1946) ; Captain James R. 
Stockman, USMC, The Battle for Tarawa, His- 
torical Division, Division of Public Information, 
Headquarters, U.S. Marine Corps (Washington, 



ners in Washington. MacArthur, having 
left the Philippines in early 1942, was de- 
termined to return as quickly and with as 
strong a force as possible. He was also 
determined that the major role in this 
undertaking should be assigned to the 
forces under his command and that the 
main approach to the Philippines should 
be made through his own Southwest Pa- 
cific theater. 

In early 1943 MacArthur's immediate 
concern was with current operations lead- 
ing up to the eventual capture of Ra- 
baul. 27 At the same time, his staff was 
preparing a long-range plan (Reno I) for 
a return to the Philippines, As first set 
forth in February 1943, this plan called 
for the progressive seizure in New Guinea 
of Hansa Bay, Hollandia, Geelvink Bay, 
and the Vogelkop Peninsula. With the 
north coast of New Guinea under control, 
Southwest Pacific forces would then ad- 
vance north to Halmahera or to the Cele- 
bes before the final jump into the south- 
ern Philippines. 28 

This schedule of operations expressed 
perfectly MacArthur's fundamental strate- 
gic ideas. The Philippines could best be 
approached by a series of amphibious 
jumps along the entire northern coast of 
New Guinea, each so distanced as to per- 
mit full cover by land-based aviation. A 
similar move into Halmahera or to the 
Celebes would bring him to the threshold 
of the Philippines. Then, with his eastern 
flank secured by previous capture of the 
Palaus and his western flank possibly pro- 
tected by the occupation of islands in the 

27 See John Miller, jr., CARTWHEEL: The 
Reduction of Rabaut, UNITED STATES ARMY 
IN WORLD WAR II (Washington, 1959). 

28 See Robert Ross Smith, The Approach to 
the Philippines, UNITED STATES ARMY IN 
WORLD WAR II (Washington, 1953), pp. 3-4. 

Arafura Sea, he would be fully prepared 
to make good his promise to return. 

Preoccupied as he was with his own 
theater, MacArthur could only view with 
alarm the growing pressure for an advance 
through the Central Pacific. The Joint 
Chiefs of Staff and the Combined Chiefs 
of Staff notwithstanding, he strongly op- 
posed an invasion of the Marshall Islands. 
Even after the Combined Chiefs had ap- 
proved and authorized the Marshalls oper- 
ation, he radioed General Marshall: 

From a broad strategic viewpoint I am 
convinced that the best course of offensive 
action in the Pacific is a movement from Aus- 
tralia through New Guinea to Mindanao. 
This movement can be supported by land 
based aircraft which is utterly essential and 
will immediately cut the enemy lines from 
Japan to his conquered territory to the south- 
ward. By contrast a movement through the 
mandated islands will be a series of amphibi- 
ous attacks with the support of carrier based 
aircraft against objectives defended by Naval 
units and ground troops supported by land 
based aviation. Midway stands as an exam- 
ple of the hazards of such operations. More- 
over no vital strategic objective is reached 
until the series of amphibious frontal attacks 
succeed in reaching Mindanao. The factors 
upon which the old Orange plan were based 
have been greatly altered by the hostile con- 
quest of Malaya and the Netherlands East 
Indies and by the availability of Australia as 
a base . . . . Z9 

This protest may have helped persuade 
the Joint Chiefs to postpone the Marshalls 
operation until the Gilberts were taken, so 
but it did nothing to sway the majority of 
Washington planners from their determi- 
nation to attack through the Central Pa- 
cific. Two months later, in August at the 

2!) Rad, MacArthur to Marshall, C3303, ao Jun 
43, CM-IN 13149- 

30 See Crowl and Love, Gilberts and Marshalls, 
pp. 1 8-* 1. 



Quebec meeting of the Combined Chiefs 
of Staff (Quadrant), it was decided, 
against General MacArthur's previous ad- 
vice, to bypass Rabaul. 31 These decisions 
on the part of higher authority did nothing 
to dissuade the Southwest Pacific com- 
mander from continuing his opposition to 
an extension of the Central Pacific drive. 
Specifically, he was strongly set against an 
invasion of the Marianas. 32 

Enter the Army Air Forces 

In the autumn of 1943 a new factor 
entered into the strategic picture of the 
war against Japan — a factor that was to 
have an important bearing on the decision 
to invade the Marianas. The Army Air 
Forces announced the imminent appear- 
ance of a very long range bomber' — the 

An experimental model of the plane was 
first flown in September 1942, but it took 
about another year to iron out the "bugs" 
and make arrangements for quantity pro- 
duction. From the point of view of stratc- 

al JCS 446; Min, 101st mtg JCS; CCS 301/3; 
and CCS 319/5. All cited n. 24. MacArthur's op- 
position to bypassing Rabaul was strongly put in 
a radio to Marshall, 04183, 23 Jul 43, CM-IN 

S2 General MacArthur reiterated his opposition 
to the invasion of the Marianas as late as 1955. 
He stated, "The 'Central Pacific Concept' . . . 
lost its validity when it was abandoned in favor 
of a 'Europe First' policy as the Japanese actual- 
ly struck the Philippines. That was the. time which 
presented the golden opportunity, both in strategy 
and logic, for a Central Pacific drive by our com- 
bined fleets aimed at engaging and destroying 
Japan's naval power on the Pacific .... Having 
missed this initial opportunity, the belated Cen- 
tral Pacific drive toward the Marianas in July 
[sic] T944 could at best produce local tactical 
successes without bringing to bear any decisive 
influence upon the course of the war." Ltr, Mac- 
Arthur to Maj Gen A. C. Smith, Chief of Mil 
Hist, Dept of the Army, J Jan 55, OCMH. 

gic bombing, the outstanding characteris- 
tic of this four-engine plane was that with 
a bomb load of four tons it had an esti- 
mated range of approximately 3,500 miles. 
In effect, once the B-29 was produced in 
sufficient quantity, mass bombing raids 
could be conducted from friendly air bases 
against enemy targets located as much as 
1,750 miles away, although for optimum 
efficiency and safety a 1,500-mile radius 
was usually used as a basis for calcula- 

As 1943 drew to a close, it became ap- 
parent that the B-29's would not be off 
the production line in sufficient number 
in time to play a significant role in the 
preinvasion bombardment of Europe, and 
that in any case the B-17's and B-24's 
already assigned to the European theater 
were adequate for the job there. 34 The 
question then arose as to how the B-29's 
could best be employed against Japan. To 
which of the various theaters of operations 
in the Pacific and Far East should the 
bulk of the bombers be assigned? Three 
possibilities suggested themselves: Austra- 
lia, China, and the Marianas. 

From the Southwest Pacific came ur- 
gent representations by Lt. Gen. George 
C. Kenney of the Fifth Air Force that first 
priority in the allocation of the new bomb- 
ers be assigned to his command. He ar- 
gued that the best way of using the B-29's 
against Japan was to knock out the pe- 

33 For an excellent treatment of the develop- 
ment of the B-29, sec Wesley Frank Craven and 
James Lea Cate, cds., The Army Air Forces in 
World War II, Vol. V, The Pacific: Matterhorn 
to Nagasaki, June 1944 to August 1945 (here- 
after cited as Craven and Cate, AAF V) (Chi- 
cago, The University of Chicago Press, 1953), 
Ch. I. 

a '' H. H. Arnold, General of the Air Force, 
Global Mission (New York, Harper & Brothers, 
1949), P. 477. 



troleum industry in the Netherlands Indies 
from bases in Australia. In Washington, 
however, the Air Forces chief, General 
Arnold, and his staff had other ideas, and 
Kenney's suggestion was rejected. 80 

At the Quebec conference of August 
1943 General Arnold presented to the 
Combined Chiefs his "Air Plan for the 
Defeat of Japan." 36 Arnold estimated that 
by October 1944 ten B-29 groups of 
twenty-eight planes each might be avail- 
able for employment against Japan. It was 
assumed that by that time no Pacific is- 
land within fifteen hundred miles of the 
Japanese main island of Honshu would 
have been captured. Therefore he pro- 
posed to build a chain of airfields north 
and south of Changsha in China, all of 
which would be within the required range 
of most of Japan's war industries. Since 
the Air Forces high command refused to 
believe either that a port on the east coast 
of China could be captured in time to 
supply these operations or that the Burma 
Road could be opened, it concluded that 
logistical support of any airfield built in 
China must come from India, flown over 
the Hump in B-24's. 37 

Air Forces planners were none too 
happy about basing their new bombers in 
China, partly because of the logistical dif- 
ficulties involved and partly because they 
were skeptical of the ability of the Chinese 
to hold the fields against the Japanese. 38 
Hence, after the conclusion of the Quebec 

30 Graven and Gate, AAF V, pp. 12, 316; 
George C. Kenney, General Kenney Reports, A 
Personal History of the Pacific War (New York, 
Duell, Sloan and Pearce, 1 949), pp. 37B, 419, 426. 

3(5 CCS 323, 20 Aug 43, title: Air Plan for 
the Defeat of Japan. 

37 Craven and Cate, AAF V, pp. 17-18. 
3H Arnold, Global Mission, p. 477. 

conference, they urged that the Marianas 
be seized and that D Day for the opera- 
tion be advanced to mid- 1944 by neutral- 
izing and bypassing intervening Pacific is- 
lands. 39 The Air Forces planners argued 
strongly that "plans for the acceleration 
of the defeat of Japan would place em- 
phasis upon the seizure of the Marianas 
at the earliest possible date, with the es- 
tablishment of heavy bomber bases as the 
primary mission." 40 The Marianas, it 
will be recalled, were about 1,270 miles 
from Tokyo, well within the estimated op- 
timum 1,500-mile cruising radius of the 

Cairo Conference 

At last Admiral King had a powerful 
ally in his persistent campaign for an in- 
vasion of the Marianas. At the meeting of 
the President and Prime Minister with the 
Combined Chiefs in Cairo in December 
1943 (Sextant), the joint Navy-Air 
Forces efforts bore fruit. Among the oper- 
ations submitted to and approved by 
President Roosevelt and Prime Minister 
Churchill was the "seizure of Guam and 
the Japanese Marianas," tentatively set for 
1 October 1944. 

This operation was to follow the capture 
of important objectives in the Marshalls 
in January, Ponape in the Carolines in 
May, and Truk in July. Meanwhile, Gen- 
eral MacArthur was scheduled to seize 
Kavieng on New Ireland, Manus Island 
in the Admiralties, and Hansa Bay on the 
northeast coast of New Guinea, and then 

■'" Craven and Gate, AAF V, p. 19. 

40 JPS 288, 4 Oct 43, sub: Plans for Defeat of 
Japan Within 12 Months After Defeat of Ger- 
many, CCS 38 r, Japan (8-25-42), Sec. 7. 



move on to the tip of Vogelkop Peninsula 
by August 1 944. 41 

The Combined Chiefs chose this occa- 
sion to again endorse the Central Pacific 
route as part of a two-pronged drive to 
Japan. In presenting their "Overall Plan 
for the Defeat of Japan," they stated that 
their ultimate aim was "to obtain objec- 
tives from which we can conduct intensive 
air bombardment and establish a sea and 
air blockade against Japan, and from 
which to invade Japan proper if this 
should prove necessary." This would ne- 
cessitate one advance along the New 
Guinea - Netherlands Indies - Philippines 
axis and another through the Central Pa- 
cific in time for a major assault in the 
area of Formosa-Luzon-China by the 
spring of 1945. The two lines of advance 
were to be "mutually supporting," but 
should there be conflicts, "due weight 
should be accorded to the fact that opera- 
tions in the Central Pacific promise at this 
time a more rapid advance toward Japan 
and her vital lines of communication; the 
earlier acquisition of strategic air bases 
closer to the Japanese homeland; and, of 
greatest importance, arc more likely to 
precipitate a decisive engagement with the 
Japanese Fleet." 42 Here in a capsule was 
the rationale of the Central Pacific con- 
cept of strategy. 

Scheduling Operations 

Upon receiving word of these latest de- 
cisions of the Combined Chiefs of Staff, 
Admiral Nimitz set about preparing a 
schedule for forthcoming operations in the 

Central Pacific. A preliminary draft of his 
campaign plan, Granite, was finished by 
27 December. It tentatively outlined oper- 
ations as follows: 43 

Capture of 

(and air attack 

on Truk ) 

Mortlock (Nomoi Is. ) 

Saipan, Tinian, 
and Guam 

31 January 1944 
20 March 1944 

20 April 1944 

1 May 1944 

1 July 1944 
15 August 1944 
1 5 November 1 944 

On 13 January, Nimitz issued another 
Granite plan revising his original some- 
what. 44 Operations to seize Mortlock Is- 
land and Truk were scheduled for 1 Au- 
gust. The possibility of bypassing Truk 
was considered, and it was suggested that 
if Truk could be bypassed, the Palaus 
should be invaded by Central Pacific 
forces on 1 August. The Marianas could 
then be invaded by 1 November since cap- 
ture of the Palaus, it was assumed, would 
be a less costly and time-consuming ven- 
ture than assaulting the Japanese strong- 
hold on Truk. In any case, both the 
original plan and the revision looked to 
the Marianas as the culmination of the 
1944 campaign. 

A curious turn of events in Pacific 
planning now took place. In order better 
to co-ordinate future operations in the two 
Pacific theaters, a meeting of representa- 
tives of the Southwest Pacific and the Pa- 
cific Ocean Areas was convened at Pearl 

41 CCS 397 (Rev,), 3 Dec 43, title: Specific 43 CINCPOA Campaign Plan Granite, 37 Dec 
Opns for Defeat of Japan. 43. 

42 CCS 417, a Dec 43, title: Overall Plan for 44 CINCPOA Campaign Plan Granite, 13 Jan 
Defeat of Japan. 44. 



Harbor in the last days of January 1944. 
Present, among others, were Admiral Nim- 
itz, and from his staff, Rear Adm. Charles 
H. McMorris, Rear Adm. Forrest P. Sher- 
man, and Vice Adm. John H. Towers, 
Commander, Air Force Pacific. From the 
Southwest Pacific came Maj. Gen. Rich- 
ard K. Sutherland, Mac Arthur's chief of 
staff; General Kenney, Commander, Al- 
lied Air Forces, Southwest Pacific Area; 
and Vice Adm. Thomas C. Kinkaid, Com- 
mander, Allied Naval Forces, Southwest 
Pacific Area. 43 

Admiral Nimitz presented his revised 
Granite plan for the consideration of the 
conferees. Immediately, and from all sides, 
objections were voiced to the proposal to 
invade the Marianas. General Sutherland 
advocated pooling all available resources 
in the Pacific and concentrating upon 
operations in MacArthur's theater. "If 
Central Pacific will move against Palau as 
the next operation after the Marshalls," 
he argued, "and make available to South- 
west Pacific Area the amphibious force 
now contemplated for Truk, we can take 
all of New Guinea, the Kai and Tanim- 
bars, and Halmahera in time to join you 
in amphibious movement to Mindanao 
this year." General Kenney spoke of 
bombing Japan by B-29's based on the 
Marianas as "just a stunt." Admiral Kin- 
kaid remarked that "any talk of the Mari- 
anas for a base leaves me entirely cold." 4fi 

Even Nimitz' own staff members 
showed themselves to be less than cnthusi- 

45 Min, Pacific Conf, 27-28 Jan 44, OPD 
334.8, Case 125; Memo, Col William L, Ritchie 
for Handy, 4 Feb 44, sub; Brief of Pacific Conf, 
Held at Pearl Harbor 37-38 Jan 44, same file; 
Hayes, The War Against Japan, Sec. IV, Part II, 
Ch. XXI, pp. 6-8; Kenney, General Kenney Re- 
ports, pp. 347-49- 

40 Min, Pacific Conf, and Mewio, Ritchie for 
Handy, cited n. 45. 

astic over the Marianas, although Nimitz 
himself favored the project. Admiral Mc- 
Morris doubted if long-range bombing 
from the Marianas would cause the capit- 
ulation of Japan. Admiral Sherman ad- 
mitted that operations in the Marianas 
would be extremely costly and that when 
captured the harbors would be of limited 
usefulness to the Navy, 

When word of these proceedings reached 
Admiral King, he read them "with indig- 
nant dismay." In a stern message to Nim- 
itz he pointed out, "The idea of rolling up 
the Japanese along the New Guinea coast, 
through Halmahera and Mindanao, and 
up through the Philippines to Luzon, as 
our major strategic concept, to the ex- 
clusion of clearing our Central Pacific line 
of communications to the Philippines, is 
to mc absurd. Further, it is not in accord- 
ance with the decisions of the Joint Chiefs 
of Staff." Assuming correctly that Nimitz 
agreed with his own strategic ideas, he 
continued, "I'm afraid . . . that you have 
not . . . maintained these views sufficient- 
ly positively vis-a-vis the officers from the 
South and Southwest Pacific." 47 Admiral 
King was not one to stand idly by while 
theater staffs undermined his favorite and 
long-nourished war plan almost at the 
very moment of its fruition. 

Acceleration of Operations 

On 31 January Central Pacific forces 
attacked Kwajalein Atoll in the central 
Marshalls and, after a four-day fight by 
the 7 th Infantry Division and the 4th 
Marine Division, secured the objective. At 
the same time, Majuro in the eastern Mar- 

47 Memo, King for Nimitz, Ser. 00409, 8 Feb 
44, COMINCH file. 



shalls was occupied without a battle. 48 
The conquest of these important positions 
had been relatively easy, and the reserve 
troops intended for the operation had not 
been committed. Nimitz could thus speed 
up his plan for moving into the western 
Marshalls, and accordingly he launched, 
on order, an amphibious assault against 
Eniwetok on 17 February, two months 
ahead of schedule. At the same time he 
delivered a carrier strike against Truk. 49 

Eniwetok fell in six days before the com- 
bined assault of the 2 2d Marines and the 
1 06th Regimental Combat Team of the 
27 th Infantry Division. On 17-18 Febru- 
ary (Tokyo time) Rear Adm. Marc A. 
Mitscher's Fast Carrier Force struck the 
once mighty Japanese bastion at Truk, 
destroying at least seventy planes on the 
ground and in the air and about 200,000 
tons of merchant shipping in the harbor. 50 
The defenses of Truk were so weak as to 
lend strong support to the idea that it 
might be bypassed altogether. 

Shortly after the successful conclusion 
of these operations in the Central Pacific, 
General MacArthur found opportunity to 
step up his own schedule. The most re- 
cently approved plans called for the South- 
west Pacific Area commander to conduct 
simultaneous invasions of Kavicng on New 

48 See Crowl and Love, Gilberts and Mar- 
shall! , pp. 302-04; Lt. Col. Robert D. Hcinl, Jr., 
USMC, and Lt. Col. John A. Crown, USMC, 
The Marshalls: Increasing the Tempo, Historical 
Branch, G-3 Division, Headquarters, U.S. Marine 
Corps (Washington, 19 '54). 

49 Rad, CINCPOA to COMINCH, 022212, 2 
Feb 44, CM-IN 1855; Min, 143d mtg JCS, 3 Feb 
43; Rad, CINCPOA to COMINCH, 150749/2, 
15 Feb 44, CM-IN 10592. 

5u See Samuel Eliot Morison, History of United 
States Naval Operations in World War II, Vol. 
VII, Aleutians, Cilberts and Marshalls, June 194Z 
-April IQ44 (Boston, Little, Brown and Com- 
pany, 1 951), Ch. XVIII. 

Ireland and Manus in the Admiralties on 
1 April. 31 Then, on 23 February, an inci- 
dent took place that persuaded MacAr- 
thur that he could safely accelerate at 
least part of this plan. Planes from the 
Southwest Pacific flying over the Admiral- 
ties on that date reported no evidence of 
the enemy. The general decided to act at 
once. He dispatched elements of the 1st 
Cavalry Division to Eos Negros to conduct 
a reconnaissance in force and, when ini- 
tial resistance was discovered to be light, 
sent the rest of the division in to capture 
the entire Admiralties group. 52 

Washington Planning Conferences 
February-March ig^4 

In the light of these events, the time had 
obviously come for the Washington plan- 
ners to reconsider their schedule of opera- 
tions for both Pacific theaters for the rest 
of 1944. 

Most planners by this time agreed that 
the primary objective for the next phase 
of the war against Japan was to establish 
a lodgment somewhere in the "strategic 
triangle" represented by Luzon, Formosa, 
and the neighboring China coast. From 
there it was believed that communications 
between the Japanese homeland and the 
Netherlands Indies could be completely 
cut off, bases for the very long range 
bombers could be set up within effective 
range of the enemy's industrial centers, 
and forward bases could be established for 
the ultimate invasion of Japan, if that 
operation should prove necessary. 

51 Ms s , JCS to CINCSWPA, 23150/Z, Jan 
44, CM-IN 15765, and to CINCPOA, 231515/Z, 
Jan 44, CM-IN 15699. 

62 Sec Miller, CARTWHEEL: The Reduction 
of Rabaul. 



Beyond this point agreement ceased. 
The arguments that ensued were many 
and various, but they can be resolved into 
two main issues. First, which was the better 
route of approach to the strategic triangle, 
the Central Pacific route through the 
mandated islands or the Southwest Pacific 
route along the coast of New Guinea to 
Mindanao? Second, which of the two 
theaters would be given priority in the 
allocation of resources, especially those 
troops, ships, and aircraft that had until 
now been assigned to the South Pacific, a 
command that had practically completed 
its mission and was about to become a rear 

Early in February Admiral Sherman 
from Nimitz' staff and General Suther- 
land, MacArthur's chief of staff, appeared 
in Washington to represent their respec- 
tive commanders in the discussions that 
the Joint Chiefs of Staff and their subor- 
dinate committees were to hold. Suther- 
land argued that Reno and Granite 
plans were "relatively weak and slow of 
progress." As a substitution, he proposed 
"an advance along the general axis, New 
Guinea-Mindanao, with combined forces." 
Truk, he believed, could be bypassed and, 
with "the capture of Truk thus obviated, 
amphibious forces can be combined for an 
advance along the northern coast of New 
Guinea." This, he claimed, would enable 
United States forces to enter Mindanao as 
early as i December 1944. For such a 
drive, naval forces could be based at 
Manus Island. He did not propose to limit 
the freedom of action of the Pacific Fleet. 
Admiral Nimitz and General MacArthur 
could operate by co-operation. But, he 
added, the "Southwest Pacific Area needs 
certain naval forces for direct support of 
its operations. It is General MacArthur's 

hope that Admiral Halsey will be assigned 
as Commander, Allied Naval Forces, be- 
cause of his ability, rank, prestige, and ex- 
perience." 53 

From General MacArthur himself came 
representations of the same nature. To 
General Marshall he radioed : 

There are now large forces available in the 
Pacific which with the accretions scheduled 
for the current year would permit the execu- 
tion of an offensive which would place us in 
the Philippines in December if the forces 
were employed in effective combination. 
However, under the plan of campaign that 
has been prepared in Washington, the forces 
will be employed in two weak thrusts which 
can not attain the major strategic objective 
until several months later . . . . 54 

In the same message MacArthur in- 
sisted that the forces of the South Pacific 
should remain under his command. These 
forces, he argued, had been engaged in 
operations within his own theater since 
their advance from Guadalcanal. He could 
not continue to operate effectively with- 
out them. "I must state," he added, "that 
any reduction in the forces presently en- 
gaged in the Southwest Pacific by actual 
withdrawal of forces of any category 
would be incomprehensible." 55 

At the same meeting of the Joint Chiefs 
of Staff addressed by General Sutherland, 
Admiral Sherman attempted to explain 
Nimitz' Plan Granite. The plan, he 
pointed out, differed from that proposed 
by General MacArthur in that it "envis- 
ages occupation of Luzon at the same time 
that an attack is made in the south 
[Mindanao] and is predicated upon the 
occupation of Eniwetok about 1 May, the 

ss Min, 145th mtg JCS ; 8 Feb 44. 
51 Msg, MacArthur to CofS, Ciai7, a Feb 44, 
CM-IN 1443. 
55 Ibid. 



Carolines about i August, and the Mari- 
anas, or such other point as might be se- 
lected as the next objective, by the end of 
the year." 56 

MacArthur's chief antagonist in this 
strategic debate was neither Sherman nor 
Nimitz, but, as usual, Admiral King. King 
vigorously opposed handing MacArthur 
the naval forces of the South Pacific. Mac- 
Arthur already had the Seventh Fleet un- 
der his command, King argued. The South 
Pacific forces were operating in a separate 
area and were "primarily concerned in 
such circumstances with the probability of 
enemy forces from the Pacific Ocean Areas 
threatening the operations of both the 
South and Southwest Pacific" King could 
see no sound reason for placing them un- 
der MacArthur.'" He called attention to 
the decision of the Combined Chiefs of 
Staff at Cairo (Sextant) that the ad- 
vance in the Pacific should be along two 
axes and that in case of conflicts between 
the two, "due weight should be accorded 
to the fact that operations in the Central 
Pacific promise at this time a more rapid 
advance toward Japan and her vital line 
of communications." 38 "General MacAr- 
thur," said King, "has apparently not ac- 
cepted this decision and desires a commit- 
ment to an advance along a single axis. I 
do not think that this is a propitious time 
to change our agreed strategy." 59 

At this point in the argument, General 
Marshall suggested that, in the light of 
developments since the Cairo conference, 
the time had come for the Joint Chiefs to 
issue a new directive to both Pacific com- 

manders. Specifically, Marshall recom- 
mended that the Joint Strategic Survey 
Committee, which consisted of Vice Adm. 
Russell Willson (Navy), Lt. Gen. Stanley 
D. Embick (Army), and Maj. Gen. Muir 
S. Fairchild (Army Air Forces), be direct- 
ed to study the matter anew and report its 
views as to what geographic objectives 
should be seized, the order of their seizure, 
and what axis of advance appeared to offer 
the best chance for victory in the Pacific. 60 
Admiral King immediately concurred. fil 

To the disappointment of the advocates 
of the Southwest Pacific concept of strat- 
egy, the Joint Strategic. Survey Committee 
came forth with a statement clearly favor- 
ing King's and Nimitz' strategic plan. 62 
The committee, repeating its earlier con- 
victions, stated that the Joint Chiefs of 
Staff "should resolve the present situation 
as between these two plans by deciding 
and directing that the primary effort 
against Japan will be made through the 
Central Pacific, with operations in the 
Southwest Pacific cooperating with and 
supporting that effort." The primary ob- 
jective, said the committee, was the 
Formosa-Luzon-China triangle, and that 
objective "would seem to be more effec- 
tively supported by the Central Pacific 
concept than by the concept of the South- 
west Pacific. The former leads most direct- 
ly and most promptly to the vital Formosa, 
Luzon, China coast area. The latter after 
reaching Mindanao will require further ex- 

! « Min, 145th mtg JCS, 8 Feb 44. 

" Memo, King for CofS, 8 Feb 44, sub: 
CINCSWPAC Despatch C121702 Feb 44, OPD 
381, Case 30 1. 

58 CCS 417, cited n. 42. 

S ' J Memo cited n. 37. 

8U Memo, CofS for King, 10 Feb 44, OPD 381, 
Case 301. 

01 Memo, King for Marshall, 11 Feb 44, sub: 
1944 Opns in Pacific Theater, OPD 381, Case 301. 

"" Apparently this report was written mostly by 
Admiral Willson. General Embick, the Army rep- 
resentative, however, was in general accord. 
Memo, Gen T. T. Handy, no addressee, 19 Feb 
44, OPD 381, Case 301. 



tensive operations before reaching that 
vital area." 03 

The Joint Strategic Survey Committee 
report not only disappointed the represent- 
atives of the Southwest Pacific but also 
failed to satisfy General Marshall, He did 
not feel that the committee had sufficient- 
ly explored the problem of allotting re- 
sources between the two theaters or the 
question of how best to employ the great 
Allied superiority in land-based air. 04 He 
wanted other subcommittees of the Joint 
Chiefs, specifically the Joint Staff Planners 
and the Joint Logistics Committee, to 
study the matter further. 65 In reply to 
these proposals, Admiral King, while 
agreeing in principle that further long- 
ranee studies would be beneficial, ex- 
pressed his fear that any more delays in 
committee might kill the momentum of the 
drive now under way in the Pacific. 60 

With Admiral King pressing for imme- 
diate action either in the direction of Truk 
or straight for the Marianas or the Palaus, 
General MacArthur on 5 March came 
forward with a proposal to accelerate 
operations in his own theater. He advised 
the Joint Chiefs of Staff to omit the Hansa 
Bay operation scheduled for about 22 
April and to move instead to Hollandia, 
some three hundred miles up the coast of 
New Guinea. To accomplish this, he pro- 
posed to use not only the forces of the 
South and Southwest Pacific Areas but 

60 JCS 713, 16 Feb 44, title: Strategy in the 

64 Memo, CofS for Leahy and King, 24 Feb 
44, OPD 381, Case 301. 

es Memo, Marshall for COMINCH and CNO, 
1 Mar 44, sub: Your Memo of 24 Feb on Pro- 
posed Directive to CINCPOA and Memo of 27 
Feb on JCS 713, OPD 381, Case 297. 

"" Memo, King for CofS, 1 Mar 44, OPD 381, 
Case 297. 

also Central Pacific aircraft carriers and 
other shipping tentatively earmarked to 
support the Kavieng-Manus operation. 67 

These suggestions were in keeping with 
General MacArthur's latest Reno plan, 
which reached Washington within a few 
days. It proposed a four-phase program: 
first, a continued advance along the north 
coast of New Guinea through Hollandia 
to Geelvink Bay; second, establishment of 
air bases in the Arafura Sea area for stra- 
tegic bombing in the Netherlands Indies 
and to support subsequent operations into 
the Vogelkop and Halmahera; third, seiz- 
ure of the western tip of the Vogelkop and 
Halmahera; and fourth, occupation of 
Mindanao, southernmost of the Philip- 
pines, and the establishment of bases there 
for an attack upon the Formosa-Luzon- 
China coast area. eH 

Admiral Nimitz, who had meanwhile 
been summoned to Washington, was quick 
to note that these proposals, if accepted, 
would have the effect of slowing up opera- 
tions in his own theater. To the Joint 
Chiefs he argued that a retention by Mac- 
Arthur of forces from the Central Pacific 
after the capture of Kavicng and Manus 
would result in "stopping the Central Pa- 
cific Campaign, losing its momentum, de- 
ferring movement into the MARIANAS 
until the approach of the typhoon season, 
and by allowing the enemy additional time 
to strengthen his defenses in the CARO- 
LINES and MARIANAS would jeopar- 
dize our ability to reach the LUZON- 

07 Msg, MacArthur to JCS, C2473, 5 Mar 44, 
CM-IN 3318. 

6S IV, Outline Plan for Operations of 
the Southwest Pacific Area to Include the Preoc- 
cupation of the Southern Philippines, 6 Mar 44, 
CNO (WPD) file, Env 68, Case 184, NR&H. 



FORMOSA-CHINA area in early 1945 
as now planned." 68 

On the affirmative side, Nimitz sug- 
gested two alternative schedules for the 
remainder of 1944. The first contemplated 
an invasion of Truk on 15 June, the 
southern Marianas on 1 September, and 
the Palaus on 15 November. The second 
proposed bypassing Truk and going into 
the southern Marianas on 15 June, Woleai 
on 15 July, Yap on 1 September, and the 
Palaus on 1 November. On reconsidera- 
tion, Nimitz decided that, if the second 
schedule were accepted, the capture of 
Yap could be deferred until the Palaus 
had been taken and that a fleet harbor 
could be established in Ulithi. This would 
advance the target date for the Palaus to 
1 October. 70 

In the end, the Joint Chiefs of Staff ac- 
cepted neither MacArthur's nor Nimitz' 
schedules in toto. Nor did they accept 
without change the final conclusion of the 
Joint Strategic Survey Committee that to 
seize the desired objective in the Formosa- 
Luzon-China coast area, "a fundamental 
strategic prerequisite is our control of the 
Marianas, Carolines, Palau IPacific] 
Ocean area." 71 At the insistence of Gen- 
eral Marshall, Mindanao was added to the 
vital intermediate objectives that United 
States forces must capture before proceed- 
ing on to the strategic triangle. 72 

" 9 Memo, CINCPAC, no addressee, 7 Mar 44, 
sub: Sequence and Timing of Opns, Central Pa- 
cific Campaign, with JCS Info Memo 200, same 
date, CCS 381 POA (6-10-43), Sec. a. 

70 Memo, Nimitz for King, 8 Mar 44, sub; 
Sequence and Timing of Opns, Central Pacific 
Campaign, CCS 381 POA (6-10-43), Sec. 2. 

71 JCS 713/3, 11 Mar 44, sub: Future Opns in 
the Pacific. 

72 Marshall's penciled notes on Draft Directive 
[JCS to MacArthur and Nimitz, 12 Mar 44], 
OPD 381, Case 301. 

Thus the directive that the Joint Chiefs 
of Staff issued to General MacArthur and 
Admiral Nimitz on 12 March represented 
in a sense a compromise between the Cen- 
tral Pacific and the Southwest Pacific con- 
cepts of strategy. It declared "that the 
most feasible approach to the Formosa- 
Luzon-China area is by way of Marianas 
-Carolines-Palau-Mindanao area, and 
that the control of the Marianas-Carolines 
^Palau area is essential to the projection 
of our forces into the former area, and 
their subsequent effective employment 
therefrom." Specifically, the Joint Chiefs 
ordered : 

1. Cancellation of the Kavieng opera- 
tion and the complete isolation of the 
Rabaul-Kavieng area with the minimum 
commitment of forces. 

2. Early completion of the occupation 
of Manus and its development as an air 
and fleet base. 

3. Occupation of Hollandia by Mac- 
Arthur, target date 1 5 April 1 944. 

4. Establishment of control of the Mari- 
anas-Carolines-Palau area by Nimitz' 
forces by neutralizing Truk; by occupying 
the southern Marianas, target date 15 
June 1944; and by occupying the Palaus, 
target date 1 5 September 1 944, 

5. Occupation of Mindanao by Mac- 
Arthur's forces supported by the Pacific 
fleet, target date 15 November 1944. 

6. Occupation of Formosa, target date 
15 February 1945, or occupation of Luzon 
if necessary, target date 15 February 

1 945;™ 

With this directive in hand, Admiral 

Nimitz and his subordinates could at last 

73 Msg, JCS to MacArthur and Richardson for 
Nimitz, 1 a Mar 44, CM-OUT 5137. 


prepare their tactical plans in detail. The largest amphibious operation yet to be 

southern Marianas would be assaulted and undertaken in the Pacific was about to 

the target date was to be 15 June. The get under way. 

The Marianas 

In History 

The islands thus chosen as the next 
point of American amphibious assault in 
the Central Pacific had had a long and 
not altogether happy experience as minor 
pawns in the international rivalries of 
great powers. Magellan discovered them in 
1 52 1 in the course of his famous first voy- 
age around the world. Struck by the sail- 
ing powers of the native boats and by the 
similarity of their rigging to that of the 
small craft that abounded in his own Med- 
iterranean, the navigator labeled his discov- 
eries, "Islas dc los Velas Latinas" (Islands 
of the Lateen Sails). Other members of 
the expedition were more impressed by 
the natives' pilfering habits and according- 
ly called them "Islas de los Ladrones" (Is- 
lands of the Thieves), a name that re- 
mained in popular usage well into the 
twentieth century, even though it had long 
since been officially abandoned. Late in 
the seventeenth century the islands were 
officially renamed "Las Marianas" in 
honor of Maria Anna of Austria, wife of 
Philip IV and Queen of Spain, who took 
a personal interest in converting their in- 
habitants to the Church of Rome. 1 

The original native population of the 
Marianas, the Ghamorros, were a hardy 
race, probably of Polynesian origin, but 
not hardy enough to withstand the en- 
croachments of western civilization as rep- 
resented by Spanish traders and mission- 
aries. In spite of mass baptisms by the lat- 
ter, native resentment toward the rigid 
rule of the priesthood increased until final- 
ly, in the 1690's, armed revolution broke 
out in the islands. The result was inevit- 
able; the Spaniards with their superior 
weapons and organization overcame the 
rebels, killed large numbers and forced 
most of the remainder to flee. The center 
of the trouble was on Guam, and from 
there many of the remaining Chamorros 
fled to the northern Marianas, where they 
were relentlessly pursued and persecuted. 
Others finally escaped south to the Caro- 
lines to become intermingled with the 
Kanaka population of those islands. 

Spanish control, which had become pro- 
gressively weaker during the nineteenth 
century, was finally severed completely in 
1898-99. Imperial Germany was the first 
of the Western Powers to challenge Span- 

1 Historical data presented here are derived 
from the following sources: R. W. Robsori, comp., 
The Pacific Islands Year Book, 194s (Sydney, 
Australia, Pacific Publications, Limited, 1942); 
R. W. Robson, comp., The Pacific Islands Hand- 

book, 1944 (New York, The Macmillan Company, 
1946); Laura Thompson, Guam and Its People 
(Princeton, N.J., Princeton University Press, 
1947) ; OCNO Div of Naval Intel, ONI 99 Stra- 
tegic Study of Guam, 1 Feb 44 (hereafter cited as 
ONI 99). 



ish hegemony in Micronesia. 2 In August 
1885 the Germans hoisted their flag over 
Yap in the western Carolines and laid 
claim to much of the surrounding terri- 
tory. After violent Spanish protest, the 
dispute was submitted to papal arbitration 
and Spain's sovereign rights were con- 
firmed, although Germany obtained im- 
portant commercial concessions at the 
same time. Negotiations for transfer be- 
tween the two countries continued in the 
i8go's, and in 1899 Spain ceded all of 
her possessions in the Carolines, the Mar- 
shalls, and the Marianas to the German 
Empire for the sum of about $4,000,000. 
Meanwhile, a small American naval expe- 
ditionary force had seized Apra Harbor 
on Guam, and at the termination of the 
Spanish-American war that island was 
ceded to the United States. 

Another radical change in the disposi- 
tion of the entire Central Pacific area oc- 
curred as the result of World War I. 
Japan, having emerged successfully from 
her recent war against Russia and being 
anxious to expand her commercial and 
military influence throughout the Orient 
and its adjacent waters, was quick to seize 
the opportunity of a European war to 
realize her own imperialist ambitions. As 
one of the Allied Powers, Japan contribu- 
ted her share to the ultimate downfall of 
the German Empire by seizing the Caro- 
lines, the Marshalls, the Palaus, and the 
Marianas (except, of course, Guam) in 
the first year of the war. Then, under the 
Treaty of Versailles, all former German 
possessions north of the equator were man- 
dated to Japan, although they theoretically 

2 Micronesia is the name applied to the islands 
of the western Pacific Ocean including the Mari- 
anas, the Carolines, the Marshalls, and the Gil- 

still remained under League of Nations 
authority. During the i92o's and 1930's 
the Japanese proceeded vigorously to col- 
onize these new holdings and to exploit 
their economic resources for the benefit of 
Japanese economy. After 1935, when Ja- 
pan withdrew from the League of Nations, 
no further effort was made to give even 
lip service to the idea that the islands 
were mandated territories. They became, 
to all intents and purposes, outright pos- 
sessions. 3 

Meanwhile, the United States had oc- 
cupied Guam and had converted it into a 
minor naval base. By an executive order 
of President William McKinley dated 23 
December 1898, the island was placed un- 
der control of the Navy Department, a 
naval officer was commissioned as governor 
of Guam, and the same officer was ap- 
pointed by the Secretary of the Navy as 
commandant of the naval station, which 
encompassed the entire island. Progress in 
constructing a naval base of respectable 
proportions was slow and halting. In 1939 
the Hepburn Board reported to Congress 
that the area should and could be devel- 
oped into a major naval base, but the nec- 
essary appropriations failed to pass. 

The Japanese Invasion of Guam 

At the time of the outbreak of war be- 
tween the United States and Japan, the 
American garrison on Guam consisted of 
153 marines and 271 U.S. Navy person- 
nel, supported by a force of 308 Guam- 
anians. The garrison had no artillery or 
fixed defenses and, in addition to its small 
arms, possessed only a few .30-caliber and 
.50-caliber machine guns. The local naval 
surface force consisted of three small pa- 

3 See |Ch. IVj below. 



trol craft and an old oiler. One of the 
patrol craft, the. USS Penguin, mounted 
two 3-inch antiaircraft guns and the oiler 
boasted two .50-caliber antiaircraft ma- 
chine guns. 4 

Japanese designs against this tiny force 
and the outpost that it garrisoned began 
to materialize well before the attack on 
Pearl Harbor. As early as the middle of 
October 1 94 1 , the 18th Air Unit, a small 
force of reconnaissance seaplanes based in 
the Marianas, began conducting a recon- 
naissance of Guam and during the follow- 
ing November flew frequent secret photo 
reconnaissance missions over the island at 
altitudes of 3,000 meters or higher/' At 
the same time, small Japanese vessels began 
patrolling the waters around the island, 
mostly at night, and in early December 
succeeded in landing several native Guam- 
anians friendly to their cause. 

Japanese plans for an amphibious inva- 
sion of Guam were complete by 8 Novem- 
ber. On that date Map Gen. Tomitara 
Hori, who commanded the South Seas 
Detachment, a unit of about 5,500 Army 
troops, received his orders. He was to as- 
semble his forces in the Bonins in Novem- 

1 C.'ipt G. J. McMillin, USN, to CNO, i i Sep 
4fi, sub: Surrender at Guam to ihc Japanese; 
ONI 99, p. 2; JICPOA Bull 52-44, if, Apr 44, 
Guam, p. 6; Historical Division, War Depart- 
ment, Guam: Operation.'; of the J7th Division (21 
July-ro August 1944), AMERICAN FORCES 
IN ACTION (Washington, 1946) (hereafter cited 
as AFAS, Guam) p. 17; Samuel Eliot Morison, 
History of United Stales Naval Operations in 
World War IT, Vol. Ill, The Rising Sun in the 
Pacific, 1 <D3i— April 1943 (Boston, Little, Brown 
and Company, 1948), pp. 32-34. 1 H4,— TiG. 

r> 18th Air Unit Combat. Rpt, G Opn, WDC 
16141a; Land Forces, Vol. 2, NA 1166,-j, WDC 
161013. Both these documents arc now located in 
the National Archives, Washington. WDC num- 
bers are the Washington Document Center ac- 
cession numbers; NA numbers are the National 
Archives accession numbers. 

bcr and wait there until definite word of 
the first Japanese air attack against the 
United States had been received. Then, 
naval air units based on Saipan were to 
fly to Guam and attack U.S. ships and in- 
stallations. Meanwhile, Army troop trans- 
ports with a naval escort were to proceed 
from the Bonins to Guam, where landing 
operations would commence in the early 
morning of 10 December, In addition to 
the Army troops assigned, the. Navy was 
to contribute a special landing force of 
about 400 men drawn from the 5th De- 
fense Force stationed on Saipan. '' 

The planes assigned to softening up the 
target came to about twenty in number, 
including the iBth Air Unit, which had 
been busy on reconnaissance missions over 
the island since October. At 0525 on 8 De- 
cember word came to the airmen waiting 
on Saipan that the Greater East Asia War 
had begun. The message read, "Begin at- 
tack on Guam immediately. 1 ' 7 The shoot- 
ing war in the Marianas had started. It 
would not formally end for another two 
years and nine months. 

Within three, hours of receiving this 
command, Japanese planes bore down on 
Guam and bombed the American oiler 
Barnes in the harbor. Next they turned 
their attention to the patrol boat Penguin, 
which was attempting to escape to the 
open sea, and finally they dumped their 
remaining bombs over shore targets. Pen- 
guin was sunk and Barnes damaged to 

,; Japanese Studies in World War II, 55, Oper- 
ations in the Central Pacific, pp. 1— fi, ropy in 
written Account, of the Civil and Military Situa- 
tion on Guam from the Japanese Occupation 
Unit February 1944, dated November 1943 and 
January 1944, GHQ FEG, G-a Hist Sec, Doc. 

7 18th Air Unit Combat Rpt, G Opn, WDC 



the extent that she had to be abandoned. 
A second flight appeared over the island 
at 1330 the same afternoon, the planes 
concentrating their attack on the still un- 
sunk Barnes and the cable and wireless 
stations ashore. 

The following day, Japanese aircraft 
again made morning and afternoon raids. 
Two bombs struck the Marine barracks, 
and many other hits were scored. The 
Japanese airmen reported that fire from 
tlie ground was very light and that they 
had spotted no artillery emplacements and 
no mines in any of the harbors or bays — 
only a machine gun p osition in the north- 

ern suburbs of Agana. (See Map V.) 

While Guam burned under the bombs 
of its attackers during the first two days 
of the war, the troop-laden assault vessels 
and their escorts were on the way down 
from the north. At 0900 on 4 December, 
the main elements of the convoy, carrying 
the 5,500 Army troops, moved out of 
Hahajima and headed toward Rota, 
where they were joined by the ships carry- 
ing the small naval detachment from Sai- 
pari. From Rota the force advanced in 
separate, groups to Guam, where all ar- 
rived during the first hour of 10 Decem- 
ber. Landing operations commenced at 
0230. The main force of the Army troops 
landed on the west coast between Facpi 
Point and Merizo, intending to drive 
northward along the coast to Agat. This 
plan miscarried when it was discovered 
that there was no adequate road from 
the beachhead to Agat, and the troops 
had to re-embark and land again at Facpi 
Point. The maneuver proved superfluous 
since by the time the new landing was 
completed the American garrison had al- 
ready surrendered to the smaller naval unit. 

The naval detachment, landed about 
two miles north of Agana. As it advanced 
toward the city, it flushed the machine 
gun emplacement, reported by planes the 
night before, about one kilometer from the 
city itself. The main body of the U.S. ma- 
rines had taken positions at the Marine 
rifle range on Orote Peninsula, and only 
about eighty Guamanians and a few 
Americans were in Agana when the Japa- 
nese, arrived. These few put up a stiff 
fight and twice at Agana Plaza drove the 
invaders back with rifle and machine gun 
fire before finally being overcome. After 
the skirmish on the plaza, the Japanese 
went on to occupy the naval hospital and 
the wireless station, and the naval repair 
station at Piti. Finally, at 0545, the island 
governor, Capt. C. J. Mc.Millin, USN, 
realized that resistance was useless and 
commenced negotiations for surrender. 
Soon after, the American ensign was run 
down from Government House and re- 
placed by the Imperial flag of Japan. The 
Japanese had completed the occupation of 
the Marianas chain. 8 

Description of the Island. 1 ; 

The Marianas are a chain of volcanic 
islands running in an approximately north 
-south direction from 20 32' N, 144 
54' E to 1 3 15' N, 144 43' E. From 
Farallon de Pa jams in the north to Guam 
in the south, the. chain numbers fifteen 
islands in all. Guam has an area of over 
200 square miles but about two thirds of 
the chain is little more than mountainous 
rocks and of practically no military value. 

N McMillin, Surrender at Guam to the Japa- 
nese; Morison, Rising Sun in the Pacific, pp. 184- 

Saipan 1 

Tin ion '/Q\ 

•^Aguijan I 



Hota 1 


a 10 to WMrLCS 



— I — 


F. T+Trtptf 



The four largest islands and those having 
the chief military utility are all in the 
southern half of the chain. These, from 
north to so uth, are Sa ipan, Tinian, Rota, 
and Guam. \[Map 2 ) | Of these, Rota was 
eliminated by the U.S. planners as a feasi- 
ble target because of the inaccessibility of 
most of its coast line, its inadequate har- 
bor facilities, and its general inferiority to 
the other three islands as a naval and air 
base for future operations against Japan. 

Saipan, Tinian, and Guam lay directly 
athwart or on the near flank of the ad- 
vance of Central Pacific forces from their 
westernmost base at Eniwetok to almost 
any part of Japanese-held territory that 
might become the object of future amphi- 
bious operations — the Philippines, Formo- 
sa, the Volcano Islands, the Ryukyus, and 
Japan proper. Saipan lies about a thou- 
sand nautical miles west of Eniwetok and 
over 3,200 miles from Pearl Harbor. North- 
west, Tokyo is only about 1,260 miles dis- 
tant, and Manila Bay is 1,500 miles almost 
due west. Tinian is just three miles south 
of Saipan; Guam, approximately another 
hundred miles in a southerly direction. 

The main geographic factors that plan- 
ners of the operation had to take into con- 
sideration were climate, the location of 
beaches and the approaches thereto, to- 
pography of the terrain inland from the 
beaches, the nature of the vegetation as it 
would affect military operations, harbor 
facilities, the location and size of towns 
and cities, and the location of roads, rail- 
roads, and other transportation facilities. 
The necessary information was gleaned 
from a variety of sources- -prewar surveys 
and hydrographic charts, aerial and sub- 
marine reconnaissance missions executed 
during the war, and enemy documents 
captured in earlier operations. 

The first aerial photographs of Saipan 
and Tinian were taken on 22 and 23 Feb- 
ruary 1944 by planes flown from Marc 
Mitscher's fast carrier force (Task Force 
58). At that time, a total of twenty-five 
sorties at various heights and angles was 
flown, but because of cloud formations 
only a partial coverage of the islands was 
obtained. 9 These were supplemented by 
full photographic coverage provided by a 
flight of five Navy PB4VS, escorted by 
Army bombers, that took off from Eniwe- 
tok on 18 April for a round trip of more 
than two thousand miles and thirteen 
hours in the air, A second mission flown by 
a similar group of planes on 25 April also 
photographed Guam. 10 Guam was cov- 
ered again on 7 May by six Navy photo- 
graphic "Liberators," and the northern 
islands as well as Guam were rephoto- 
graphed on 29 May. 11 

In addition, the submarine USS Green- 
ling made a scries of sorties around all 
three islands between 2 and 29 April and 
obtained excellent photographs, which 
were made available to the expeditionary 
troops. The submarine's photographs were 
chiefly remarkable for their accuracy of 
detail on the beach approaches. 12 

Between 15 January and 10 May, Joint 
Intelligence Center, Pacific Ocean Areas 
(J1CPOA), issued a total of eight infor- 

9 Hq Expeditionary Troops, Task Force 56 Re- 
port on Foraghs. (hereafter cited as TF 56 Rpt 
Forager), Incl D, G-2 Rpt on Forager, p. 2; 
App. C, pp. 1-3. 

44, pars. 37"39. 

44, pars. 49-50; TF 56 Rpt Forager, Incl D, p. 2. 
All of these Navy photographic missions were es- 
corted by Army B-24's of the Seventh Air Force 
flown out of the Marshalls. Craven and Cate, 
AAF IV, pp. 684-87. 

44, pars. 49-50- 



mation bulletins covering the proposed is- 
land targets, 13 In addition, the Office of 
Naval Intelligence (ONI) made available 
a monograph on the geography of the Pa- 
laus and the Marianas that had been pre- 
pared in the spring of 1942. 14 More de- 
tailed was another ONI bulletin on Guam 
that was an exhaustive, although some- 
what obsolete, study based chiefly on a 
monograph completed by Marine Corps 
Schools in Quantico as early as 1937 and 
the notes and personal recollections of a 
naval officer, Comdr. R. F. Armknecht 
(CEC) (USN), who had left the island 
only a month before the Japanese captured 
it. 15 Other materials, including diagrams 
of tides, sunlight and moonlight tables, and 
captured Japanese charts, completed the 
list of information on the Marianas upon 
which plans for the landings had to be 
based. 16 

Some of this proved to be inaccurate in 
detail. The number of enemy installations 
and the size and disposition of troops were 
generally underestimated. Many terrain 
features were erroneously depicted in the 
maps made on the basis of photographic 
intelligence. On the whole, however, a 
fairly good general understanding of the 
nature of the targets could be gleaned 
from the various sources. 

Climatic conditions vary little as be- 
tween the sister islands of Saipan and 

13 JIGPOA Bull 7-44, Mariana Islands, 15 Jan 
44; Bull 29-44, Weather Survey for Carolines 
and Marianas, 26 Feb 44; Bull 31-44, Carolines 
and Marianas, Part I, Tides and Currents, Part II, 
Daylight and Dark, 4 Mar 44; Bull 34-44, Sai- 
pan, Tinian, and Rota, 10 Mar 44; Bull 53-44, 
Guam (2 vols.), 15 Apr 44; Bull 66-44, Saipan 
(Target Survey), 10 May 44; Bull 67-44, Tinian 
(Target Survey), 10 May 44; Bull 73-44, Saipan, 
Tinian, and Rota, 1 May 44. 

14 ONI 29, Palau, Mariana Islands, 1 1 May 4a. 

15 ONI 99, p. vii. 

1B TF 56 Rpt Forager, Incl D, App. B. 

Tinian in the north and Guam in the 
south. The year can be conveniently di- 
vided into two seasons: the dry or winter 
monsoon season from November through 
March, and the wet or summer monsoon 
season from April through October. Dur- 
ing most of the dry period the prevailing 
wind is from north and cast with an aver- 
age velocity of ten to fifteen knots. By 
June the wind usually shifts from east to 
south and by August and September the 
southwestern monsoon frequently occurs. 
Thus, if landings on the western beaches 
were to be considered, it was clear that 
they should be made before August in or- 
der to be certain of a lee shore. 

Average cloud cover is nearly 70 percent 
for most of the year, which added to the 
problem of achieving adequate aerial ob- 
servation and aerial reconnaissance. Gen- 
erally, however, weather conditions offered 
no serious obstacle to military operations. 
The climate is mild and healthful. Mean 
temperature on Saipan varies between a 
January maximum and minimum of 81 ° 
F. and 72 ° F., respectively, and June val- 
ues of 85 ° F. and 75 ° F. Guam, being a 
hundred miles closer to the equator, is 
somewhat warmer and more humid but 
not enough so to make any significant dif- 
ference. The summer months constitute 
the rainy season, during which the total 
precipitation is from 80 to 86 inches. Most 
rain falls in the form of showers, lasting 
from only a few minutes to two hours, but 
there are occasional steady, light rains. 
August is the wettest month — an addi- 
tional reason for concluding any proposed 
military operations before the end of July. 
Typhoons occur in the area, but not with 
as much regularity or intensity as else- 
where in the Pacific. This is a region of 
typhoon genesis and storms are usually not 



fully developed. About once every two 
years disturbances of the typhoon type hit 
the islands or pass near enough to cause 
violent winds and heavy rainfall, but even 
so the storm diameter is generally com- 
paratively small. 17 

None of the three islands offered ideal 
beach conditions for landing assault 
troops. On Saipan the land on the western 
side slopes rather gently up from the shore 
line, but the beaches themselves could be 
reached only with difficulty because of 
fringing and barrier reefs that extend for 
most of the island's length. \(See Map I.)\ 
A small gap existed in the reel just off the 
town of Charan Kanoa and a dredged 
channel led into Tanapag Harbor, but for 
most of the area the coral barrier present- 
ed a serious impediment to the landing of 
troops boated in small craft. The north 
end and east side of the island have deep 
water close inshore and are free of reefs 
except for a fringing one around Magi- 
cienne Bay on the east coast. However, 
the beaches were generally narrow and the 
shores steep, thus making landing opera- 
tions and movement inshore extremely 
hazardous. The same conditions prevailed 
for the most part on the southern end. 

Harborage facilities ranged from poor to 
mediocre. Tanapag Harbor, the principal 
anchorage area, was the only one that pro- 
vided even partial protection from all 
winds, and against heavy seas from the 
west the reef barrier on the seaward side 
offered little real safety to ships at anchor 
within. Vessels could anchor off Garapan 
and find some shelter from easterly winds 
and sea, but it was no more than an open 

^JICPOA Bull 39-44, Detailed Weather- 
Guam, pp. 1-3; Detailed Weather — Saipan and 
Tinian, pp. 1-3; Comdr Fifth Fleet Opn Plan 
Central 10-44, 12 May 44, App. I, Weather 

roadstead, and landing and unloading 
would be impossible during strong wester- 
ly winds. The only shelter against norther- 
ly and westerly winds was at Magicienne 
Bay on the southeast coast. However, the 
bay was deep and open to winds and seas 
from the southeast. 18 

Tinian was even better equipped topo- 
graphically to resist an amphibious land- 
ing. The main obstacle to invasion was an 
almost unbroken barrier of abrupt cliffs 
close to the water line and ranging from 
a few feet to over a hundred feet in height. 
Although the cliffs were not unscalable, 
especially along the northern and western 
shores, there was very little landing space 
at the water's edge for small craft, and the 
flow of supplies inland from such narrow 
beaches would be seriously handicapped. 
The cliffs were broken at various spots in 
the neighborhood of Tinian Harbor and 
again on the northwest coast and along 
the northeastern shore line, but none of 
these loc ations offered ideal landing con- 
ditions. " |(Sflf Map j77TI 

Guam, like the other islands of the Ma- 
rianas, presented to would-be invaders the 
combined hazards of reefs and shore-line 
cliffs. The northern half of the island was 
virtually inaccessible to amphibious troops 
because of the reef, the surf, and the sheer 
cliffs rising from the beaches. Although 
there were some possible landing points on 
the southeast and southern coasts, the most 
practicable beaches were south of Orote 
Peninsula and north of the naval station 
at Piti, both on the southwest coast. These 
were obstructed by barrier reefs and, in 
the north, by low-level cliffs, but were at 
least negotiable by amphibian craft and 

18 JICPOA Bull 73-44, PP- 6, 1 1 ; Bull 34-44, 
p. 24; Bull 7-44, pp. 53-56. 

1B JICPOA Bull 7-44, pp. 43-49; Bull 34-44, 
p. 34; Bull 73-44, p- ao. 



vehicles. The only good harbor was at Port 
Apra, north of Orotc Peninsula. The port 
was protected by the peninsula proper, by 
Cabras Island on the north, and by a 
breakwater extending from the island, and 
was considered safe from wind and sea 
from almost any direction except during 
the typhoon season. 20 

Although terrain features inland from 
the shore line on each of the islands were 
in some ways unique, the similarities were 
more marked than the differences. The 
most distinct characteristics of all the Ma- 
rianas, at least to seasoned troops, were 
to be their novelty and variety. Heretofore, 
practically all operations of American 
forces in the Pacific had been confined to 
tropical jungle or to coral atolls where 
maneuver was limited either by the densc- 
ness of the vegetation or by the smallness 
of the area to be seized. Now American 
troops were to find themselves maneuver- 
ing over moderately large land masses with 
highly varied topography and vegetation. 
They would have to fight their way 
through mangrove swamps and fields of 
lashing sword grass and sugar cane; 
through fairly sizable towns where the 
enemy might have to be routed by house- 
to-house fighting; and, finally and worst of 
all, up and down precipitous volcanic 
mountains pocked with caves and creased 
with endless ridges, escarpments, and other 
natural concealments that the enemy could 
be expected to exploit to the fullest. 

The dominating physical feature of Sai- 
pan is Mount Tapotchau, which rises to 
an elevation of 1,554 ^ eet near tne center 
of the island. 21 Between this summit and 
Mount Marpi, located at the north end of 

20 JICPOA Bull 7-44, pp. 9-12; Bull 52-44, 
pp. 44-51. 

21 Unless otherwise stated, the elevations in the 
text of this volume are in feet. 

the island and having an elevation of 832 
feet, there is a ridge over seven miles long 
with peaks ranging between 720 and 934 
feet. To the west of this ridge there is a 
fairly level coastal area in the region of 
Tanapag Harbor, but to the cast the slope 
ends abruptly in steep coastal cliffs. South 
and southeast of Tapotchau, the backbone 
ridge slopes off into the plateau forming 
Kagman Peninsula on the eastern and the 
southern third of the island. In the south- 
ern area Mount Kagman, on the east, and 
Mount Nafutan, on the southern peninsula 
of the same name, form conspicuous head- 
lands. Approximately 70 percent of the 
island's 85 square miles was under sugar 
cultivation at the time of the invasion. 
Cane can present a serious obstacle to 
troops, impeding movement by foot, limit- 
ing fields of fire, and offering excellent 
concealment for the defenders. Another 
obstacle to attacking troops, especially 
from west coast beaches, was an extensive 
marsh that was inland from the town of 
Charan Kanoa and centered on fresh- 
water Lake Susupe. Saipan is better than 
fourteen miles along its north-south axis 
and about six miles across at its widest 
point. 22 

Tinian is somewhat smaller, being about 
10.5 miles in length and 5.5 miles in width 
at its widest point. Its terrain is also con- 
siderably less difficult for military opera- 
tions than is that of its sister island. It is 
basically a broad elevated limestone pla- 
teau, the highest point of which is Mount 
Lasso (564 feet) in the north-central part. 
Most of the southern end of the island 
is of rolling hills and flatlands except for 
the southernmost tip, which is marked by 
precipitous cliffs and ravines. At the time 

22 JIGPOA Bull 7-44, PP. 53-54; Bull 73-44, 
p. 6. 



of the invasion about 90 percent was in 
sugar cane and the remainder, along the 
rocky slopes between the broad terraces 
and along the coasts, was covered with 
dense vegetation. 23 

Guam, the largest of the Marianas, is 
32 miles long and from 4 to 8 miles wide, 
with a total surface of over 200 square 
miles. The northern half is a large coral 
limestone plateau sloping upward from 
the lowlands in the middle of the island 
between the town of Agana on the west 
coast and Port Pago on the east. It is 
broken in the south by Mount Barrigada 
and in the east-central section by Mount 
Santa Rosa, which rises to an elevation of 
870 feet. The southern half of the island 
is characterized by broken mountain coun- 
try. Inland from Apra Harbor on the west 
coast below Agana lie the mountain peaks 
of Chachao, Alutom, and Tenjo, all above 
a thousand feet in elevation. South of these 
peaks and opposite Agat Bay, the ridge 
slopes down to a 400-foot saddle and then 
rises again at the southern end of the is- 
land to elevations of over 1,200 feet. On 
the westward side the range slopes steeply 
to low foothills and narrow belts of low- 
lands. On the east the slopes are more 

gradual, ending in a plateau about 400 
feet high stretching to the coastal high- 
lands, which rise in steep bluffs from a 
narrow coastal flat. Five streams pierce 
the plateau in their eastward course. The 
highlands of the southern half of Guam 
are largely volcanic rock except for Orote 
Peninsula and Gabras Island on the west 
coast and the east coastal regions, all of 
which consist of coral limestone. Most of 
southern Guam, volcanic in origin, was 
covered with breast-high sword grass, 
and sparse scrub growth. In the coral 
limestone regions were found weeds, trail- 
ing vines, and tropical vegetation thick 
enough to make even roads impassible if 
not cleared continuously. In addition to 
these undesirable features, the coastal 
areas abounded in marshes and lowlands 
that were usually cultivated as rice pad- 
dies and that were either deep in mud or 
covered with cogon grass and weeds reach- 
ing heights of fifteen to twenty feet. In 
short, Guam even more than the other 
islands of the Marianas contained a super- 
abundance of terrain difficulties to harass 
and impede the progress of attacking 
troops. 24 

23 JICPOA Bull 73-44, p. 20. 

24 JICPOA Bull 7-44, pp. 8-10; AFAS, Guam, 
pp. 1 4-1 5. 


Planning the Invasion 

Organization and Composition 
of the Attack Force 

For the Marianas, as in the case of all 
operations in the Pacific outside of General 
MacArthur's jurisdiction, Admiral Nirmtz 
retained over-all command of the cam- 
paign. Under him in the chain of command 
was Vice Adm. Raymond A. Spruance, 
commander of the Fifth Fleet, and under 
him Vice Adm. Richmond Kelly Turner, 
who was to command the Joint Ex- 
peditionary Force ( Task Force 51), was 
charged with the actual job of taking the 
islands. Turner wore a second hat. Until 
15 July 1944, he was also in command of 
the Northern Attack Force (Task Force 
52), which was made up of all the am- 
phibious elements assigned to the attack on 
Saipan and Tinian, and which was one of 
the two component parts of Task Force 5 1 . 
Its equivalent for Guam was designated 
Southern Attack Force (Task Force 53) 
and was commanded by Rear Adm. Rich- 
ard L. Conolly. Vice Adm. Marc A. 
Mitscher's Fast Carrier Task Force (Task 
Force 58 ) and Vice Adm. Charles A. Lock- 
wood's Submarine Force, Pacific Fleet 
(Task Force 17), were assigned supporting 
missions according to their appropriate 
capacities. The former operated as part of 
the Fifth Fleet and the latter directly under 
Admiral Nimitz. YChart 1)\ 

Tactical command of all troops ashore 
for the Marianas operation devolved upon 
Lt. Gen. Holland M. Smith, USMC. Gen- 
eral Smith was perhaps as well grounded 
in the fundamentals of amphibious tech- 
niques as any general officer in either the 
Army or the Marine Corps at that time. 
In 1 94 1 and 1942 he had supervised the 
training of the 1st Infantry Division and 
the 1 st Marine Division in basic landing 
problems on the U.S. east coast. At that 
time amphibious warfare was still some- 
thing of a novelty, and United States 
forces were generally innocent of the fun- 
damentals of launching an assault of sea- 
borne troops against a hostile shore. 

For at least two decades before the 
outbreak of World War II, it is true, the 
Marine Corps had slowly been piecing to- 
gether a workable body of amphibious 
doctrine, and after 1934, in conjunction 
with the U.S. Navy, it had conducted 
yearly landing exercises, chiefly on the is- 
land of Culcbra, Puerto Rico. These had 
been valuable, indeed indispensable, ex- 
periments. They were in no small measure 
responsible for the ability of American 
troops to invade the beaches of Africa, 
Europe, and countless Pacific islands. 

Not until after the fall of France did the 
United States commence to prepare in 
earnest for large-scale amphibious landings. 
In February of 1941 General Smith and 

Chart 1 — Task Organization for Major Commands for Attack on Saipan and Tinian 

Tosh Faitf '50 

FiJlh n«*T 

Ann, R, ASpfuOnCt 




Tqik Fare* 5fl 

folk Force 51 

loinl E.*p4di1it^oiy Fnrct 

Vic* Adm R, K. limn 

Tar* Fore* ST 

Vic* AoWi. K 

\ A MlHtW 

(Land-BbHd Aircroh) 
Vie Adm. 1 H H«( ( . 

Ta»k F arc* So 
Expeditionary Troops 

Lt Grfl H M. Smilh, USMC 

HHJ ': 

Northern troopi and Landing Farce 
Ll G*n K. M Smith, USMC 

(Mai Gen. H Schmidt, USMC)' 


Tail Farce 5! 

Northern A Ha cli force 

V-ct Adm B. 

K- turtle! 

(Sea. Adm H 

w H,\\y 

V Amphibia in. Caret Traopr 
Li Gin H M Smith, USMC 

I Ma, Gm. H Schmidl. USMC)' 

Mai Gin T E Wotion, USMC 

41 h r\larinc Division 

Moj. 6*n. Hany Schmidt, USMC 

(Mai Gin, C 8 Coin, USMC)'' 

XXIV Cotpi Artillery 

fci*. Gen, A. R Horp,,, USA 

STrh Inbnlrv Diviilon ■ 

Moj. G*n. ft C SmHh, USA ' 

iMaj. Gen. S. lom>an> USA)' 
Ma, Gen. G. W Giiner) USA' 

• Aitumed command ol NiLF onll July whan h« 
wen relieved d1 command ol the Mb Marine Diviwon 

b A&umtd commgnot ol 4th Matin* Drviiion on It 

'tTtfi fnfonlty Division rtteOied dam Commander 
foik Forte SI lo NTLF on 16 June. 

J Relieved ol command ot !7lr> Infantry Diviiion 
!4 June. 

• Commanded jllfc Infantry Dinner, 24 to 58 Jtrni. 
i Attyrnrd command ol ftTrh Infantry Division SS 


• Amumcd command ol Northern Attack Fotc* 1 S 
July fat the anertk en tin ion. 

folk Group SS, 2 

t Second in Command} 
!«, Adm. H. W Hill 

Talk Group 5S.3 
TronipoH Group "Able" 
Cop) H. B Knowltr., USN 

Task Grouo 5S.4 
Transport Group "Baker*' 

Copt D. W Loofflii, USN 

Talk Group J1 S 
EaiTrr* Landing Group 
Com*. C J MeWhinnLt, USN 

Task Group 59 It 
Traniporl Screen 

Cap*. It. i Libby, USN 

Toik Group 5S.5 

Ttoctoi Flotilla 

Capi A. jt Robertson, USN 

Tatk Group S1 1 T 

Fir* Support Group On* 

Rear Adm. J. 6 Oldendorl 

Task Group H_1Q 
Firt Support Group Two 
R«r Adm. W, L Ainswotlh 

Toik Group 33.14 

Car pier Support Group One 

R*ot Adm, G. F. Boean 

Tatk Group Si .1 1 

Corner Support Group Two 

Rior Adm. H, B, Soflaoc 

Toik Group SS 13 

Minesweeping, and Hydrographic Survey 

Comdr R S Moor., LtSNR 

Toik Group 51 1 

Joint Errptrditiorrory Forcu Rttr-'r 

Sltrt Irtfonlry DivrHon' 

Moj urn R C Smiih USA 1 

(hnoi G*rt- S- Jprman, USA)* 

(Moj. G*n, G W Grmtr, USA)' 

Toik Groupj 51 1 Id 51 T 

DElrrma and Ga"iton &aupl 

h^oi. Gen. S. Jotman, USA (Saipan)" 

Mai Gen. J L. LfoderNI, USA (Tirwi] 

Toik Group 51 .8 
General Rettrve 

77th Inrontry Drvriran 

Moj Gen. A. D, Brunt 

Command and opetalronctl contrsf 

— *■ — 4 »S- —♦■ — «*— 

Operational control only unliltraopi are estobli iked 



his staff planned and oversaw a joint 
Army-Marine Corps practice landing in the 
Culebra area. In June of the same year 
the first full two-division landing exercise 
was conducted at New River, North Car- 
olina, under their supervision. Another was 
held on a somewhat smaller scale off Lynn 
Haven Roads, Virginia, in January of 
1942. Two months later, General Smith 
was ordered to duty as Commander, Am- 
phibious Corps, Atlantic Fleet, an am- 
phibious training command, and later 
in the year he served in much the same 
capacity as Commanding General, Am- 
phibious Corps, Pacific Fleet. In Septem- 
ber 1943 he and his staff left San Diego 
for the Central Pacific, where Smith was 
to be commander of the V Amphibious 
Corps. 1 As such, he commanded the ex- 
peditionary troops that captured Tarawa 
and Makin in the Gilbert Islands and 
Kwajalein and Eniwetok in the Marshalls. 2 
Now his task was even greater. 

Holland Smith's designation for this 
operation was Commanding General, Ex- 
peditionary Troops (Task Force 56). He 
was directly responsible to Admiral Turner 
until the amphibious phase was completed. 
Like Turner, he was to play a dual role. 
As Commander, Northern Troops and 
Landing Force (Task Group 56.1 ), he per- 
sonally exercised tactical control of all 
troops ashore during the capture of Sai- 
pan. He was relieved on 12 July 1944 from 
this command (but not from command 

1 Holland M. Smith, "The Development of 
Amphibious Tactics in the U.S. Navy," U.S. 
Marine Corps Gazette (October, 1946), pp. 45- 


Holland M. Smith, Coral and Brass 

(New York, Charles Scribner's Sons, 1949), Chs. 
IV, V, VI. Jeter A. Iscly and Philip* A. Crowl, 
The U.S. Marines and Amphibious War (Prince- 
ton, N.J., Princeton University Press, 1951), Chs. 

" See Crowl and Love. Gilberts and Marshalls. 

of Expeditionary Troops) by Maj. Gen. 
Harry Schmidt, USMC, who thereafter 
performed the same role during the seizure 
of Tinian. Their counterpart on Guam was 
Maj. Gen. Roy S. Geiger, USMC, the com- 
manding general of III Amphibious Corps 
and of Southern Troops and Landing Force 
(Task Group 56.2). Although in this ca- 
pacity Geiger and Smith held parallel com- 
mands, the former was subordinate to the 
latter as Commanding General, Expedi- 
tionary Troops. 3 

The command relationships among Gen- 
eral Smith, his naval superiors, and his 
Marine and Army subordinates, although 
resembling in complexity the hierarchy of 
saints, can be reduced to fairly simple 
terms. In effect Admiral Spruance enjoyed, 
by delegation from Admiral Nimitz, su- 
preme command of the operation. He re- 
tained operational command throughout 
and upon him devolved the responsibility 
of determining when the capture and occu- 
pation phase of each island had been 
completed. Tactical command during the 
amphibious phases of the operation was 
placed in the hands of Admiral Turner, 
who exercised it directly at Saipan and 
through Rear Adm, Harry W- Hill on Tin- 
ian and Admiral Conolly on Guam. 

The completion of the amphibious phase 
was determined in each instance by the 
landing force commander — whenever he 
decided that the situation warranted it, he 
was to establish his command ashore. 
Thereafter, all tactical decisions regarding 
the disposition of troops would be made 
by him. On Saipan the landing force com- 
mander was Holland Smith, on Tinian 
Harry Schmidt, and on Guam Roy S. Gei- 
ger. On all three islands, however, "overall 

3 TF 56 Rpt Foragf.r, pp. 7-8. 



Toy Naval Commanders in the Marianas Campaign. From the left: Vice 
Adm. Raymond A. Spruance, commander of the U.S. Fifth Fleet; Fleet Admiral 
Ernest J. King, Chief of Naval Operations and Commander in Chief, V ,S. Navy; 
and Admiral Chester A. Nimitz, Commander in Chief, U.S. Pacific Fleet and 
Pacific Ocean Areas. 

troop command" was retained by Gen- 
eral Smith as Commander, Expeditionary 
Troops. 4 

There were 105,859 assault troops as- 
signed to capture the three islands; 66,779 
were allocated to Saipan and Tinian and 
the remaining 39,080 to Guam. The bulk 
of the force was made up of two reinforced 
Army divisions, three reinforced Marine 
divisions, and a provisional Marine bri- 
gade consisting of two regimental combat 
teams. 6 

1 Ibid., Incl A, Opn Plan 3-44, Annex I, 
Change 1, 5 May 44. 

5 Commander Joint Expeditionary Force, Mar- 
ianas (Task Force 51), Report of Amphibious Op- 

The landing on Saipan was to be made 
by the 2d and 4th Marine Divisions, with 
the 27th Infantry Division in reserve. 6 All 
three of these organizations had seen pre- 
vious action in the Pacific. The 2d Marine 
Division was activated in San Diego on i 
February 1941. One regiment (2d Ma- 
rines) took part in the initial attack on 
Guadalcanal on 7 August 1942, and the 
remaining two entered that campaign in 

erations for the Capture of the Marianas Islands 
(Forager Operation), 25 Aug 44 (hereafter cited 
as TF 51 Opn Rpt Marianas), p. 6. 

,; See Appendix A, below, for breakdown of 
troops assigned to Northern Troops and Landing 



November and January, The division had 
also fought the bloody battle of Tarawa, 
losing over 3,000 casualties there, 7 An in- 
dependent Marine unit, the 1st Battalion, 
29th Marines, which was formed in the 
spring of 1944 around cadres of 2d Marine 
Division veterans of Guadalcanal and Tar- 
awa, was attached to the 3d Division for 
the Saipan operation. 8 In the Marianas, 
the division was commanded by Maj. Gen. 
Thomas E. Watson, USMG, who had pre- 
viously led the Marine and Army regiment- 
al combat teams that captured Eniwetok 
in the Marshalls. 

The 4th Marine Division was not form- 

1 Richard W, Johnston, Follow Me!: The Story 
of the Second Murine Division in World War II 
(New York, Random House, 194H), pp. 7-156. 

8 Bevan G. Cass, ed., History of the Sixth 
Marine Division (Washington, 1948), p. 11. 


ally activated until 16 August 1943, but it 
was by no means totally unseasoned. In 
early February 1944 it had captured Roi 
and Namur Islands in Kwajalein Atoll 
while troops of the 7th Infantry Division 
were taking nearby Kwajalein Island in the 
central Marshalls. 9 The 4th was to be com- 
manded at Saipan by General Schmidt, 
who had been the division's commander 
since it was first formed. When General 
Schmidt relieved Holland Smith of com- 
mand of the Northern Troops and Landing 
Force after Saipan was officially declared 
secure, he in turn was succeeded in com- 
mand of the 4th Marine Division by Maj, 
Gen. Clifton B, Gates, a veteran of Guad- 
alcanal and an alumnus of the 1st Marine 
Division. 1 " 

The 27th Infantry Division was a Na- 
tional Guard unit of New York State when 
it was called into federal service in Octo- 
ber r94o. Its three regiments, the 105th, 
106th, and 165th, had had their headquar- 
ters at Troy, Albany, and New York City, 
respectively. 11 It was the first combat di- 
vision to leave the United States for Pacific 
duty and by the war's end had spent a 
longer time overseas than any National 
Guard division in the United States Army. 
In March 1942 advance echelons arrived 
in Hawaii and for the next year and a half 
the division served as base defense force, 
first for the outer islands and then on 
Oahu after the 35th Division was sent to 

Vice Adm. Richmond Kelly Turner 

H Carl W. Froehl, ed., The Fourth Marine Di- 
vision in World War II (Washington, 1946), pp. 
16-33; Crowl and Love, Gilberts and Marshalls, 
pp. 304-08. 

10 Proehl, Fourth Marine Division, pp. 2-4. 

11 When the division was triangularizcd in 1942 
its fourth regiment, the ioSth Infantry, was trans- 
ferred to the 40th Division. Edmund G. Love, 
The syth Infantry Division in World War II 
(Washington, Infantry Journal Press, 1949), pp. 



Lt. Gen. Holland M. Smith 

Guadalcanal. In November 1943 the 165th 
Infantry, reinforced by the 3d Battalion, 
105th Infantry plus organic artillery, en- 
gineer, and service units, invaded and 
captured Makin simultaneously with the 
2d Marine Division's assault on Tarawa. 
Three months later two battalions ( 1 st and 
3d) of the 106th Infantry, plus an inde- 
pendent Marine regiment, 12 took Eniwetok 
Atoll. Thus, of the entire 27th Division 
only the 1st and 2d Battalions, 105th In- 
fantry, and 2d Battalion, 106th, which se- 
cured Majuro Atoll without battle, were 
unseasoned in atoll warfare. 13 

12 The 2 ad Marines. This unit later was incorpo- 
rated into the 1st Marine Provisional Brigade, 
which fought on Guam. The brigade was subse- 
quently expanded to the 6th Marine Division, 
which saw action on Okinawa. 

13 Love, The syth Infantry Division, pp, 18- 
ni ; Hq 27th Inf Div, Hist of 27th Inf Div From 
Induction, 15 Oct 40, to Date, 20 Mar 45, pp. 

Maj. Gen. Ralph C. Smith joined the 
27th Division as commanding general in 
November 1942. His previous wartime 
duty had been with the Military Intelli- 
gence Division (G-2) of the War De- 
partment General Staff and with the 
76th Infantry Division at Fort George G. 
Meade, 14 His primary job for the next 
year was to supervise training of the di- 
vision for forthcoming operations. His own 
initiation into Pacific warfare came at 
Makin, where he exercised tactical com- 
mand over the reinforced 165th Infantry 

Aside from the three reinforced infantry 
divisions, the largest single unit attached 
to Northern Troops and Landing Force for 
the Marianas operation was the XXIV 
Corps Artillery (Army). The organization 
was formally activated on 25 March 1944 
and consisted of two battalions each of 
155-mm. howitzers and 155-mm. guns. 
The nucleus of this new organization con- 
sisted of coastal artillery and field artillery 
battalions orginally assigned to the defense 
of Oahu. One battalion (145th) had par- 
ticipated in the Kwajalcin Island landing, 
but the rest were new to combat. 15 For the 
Marianas Campaign the corps artillery was 
commanded by Brig. Gen. Arthur M. 
Harper, a field artilleryman since 1920. 
Between the commencement of the war 
and his assignment to XXIV Corps, he 
had served as artillery officer of I Corps, 
of the 30th Infantry Division, and as com- 
manding general of III Corps Artillery. 1 " 

14 General Officers, Service Biographies, DRB 
AGO, Misc 301. 

13 XXIV Corps Arty Final Rpt on Forager 
Opn, Phases I and III, S-3 Rpt, pp. I-H. 

1(1 General Officers, Service Biographies, DRB 
AGO, Misc 201. 



Tactical Planning 

Headquarters, V Amphibious Corps, 
was first alerted to its forthcoming respon- 
sibilities in the Marianas on 15 January 
1944, when it received Admiral Nimitz' 
Campaign Plan Granite setting forth the 
concept and outlining a tentative schedule 
of operations for the Central Pacific area 
for the year ig44. 17 Operation Forager, 
involving the seizure, occupation, and de- 
fense of Saipan, Tinian, and Guam, was 
included as the final phase of this program. 

The first two months of 1944 witnessed 
an unexpected speed-up in Pacific, opera- 
tions. By 17 February, as already noted, 
Kwajakin Atoll had been seized, a suc- 
cessful landing had been made on Eni- 
wetok, and, most important, a fast carrier 
strike against Truk had revealed the al- 
leged impregnability of that once-powerful 
base to be a myth. 

On 1 3 March, therefore, Nimitz assigned 
highest priority to the Marianas operation. 
A week later he issued his Joint Staff Study 
for Forager to all major commanders as a 
guide for advanced planning. The study 
indicated that V Amphibious Corps, in- 
cluding the 2d and 4th Marine Divisions, 
would be mounted in the Hawaiian area 
for the initial assault on the beaches of 
Saipan with the 27 th Infantry Division in 
reserve. The III Amphibious Corps, con- 
sisting of the 3d Marine Division and the 
1 st Provisional Marine Brigade, was to be 
mounted in the Guadalcanal area for an 
invasion of Guam. The 77th Infantry Di- 
vision was to be alerted in the Hawaiian 
area for possible movement to the Mari- 

Maj. Gen. Ralph C. Smith 

anas twenty days after the initial landing 
on Saipan. The probable target day (D 
Day) for Saipan was set as 15 June. The 
date for the invasion of Guam (W Day) 
was tentatively established as 18 June. 18 
On 12 April General Holland Smith di- 
vided his V Amphibious Corps staff into 
two separate components. One, initially 
known as the Red Staff, later functioned 
as Northern Troops and Landing Force 
(Task Group 56,1 ) for the capture of Sai- 
pan and Tinian. The other portion, first 
known as Blue Staff, later served as 
Headquarters Expeditionary Troops (Task 
Force 56 ), 19 General Smith's two staffs 
were heavily augmented by U.S. Army per- 

17 See above, p. 13. Planning for the Guam and 
Tinian phase of Forager, although undertaken 
concurrently with that for Saipan, is discussed in 
detail in Chapters XIII and XV. 

18 TF 56 Rpt Forager, Incl C, G-3 Rpt, pp. 
1-3; Commander in Chief Pacific Ocean Areas, 
Forager Joint Staff Study. 

1!) Headquarters Northern Troops and Landing 
Force Report, Marianas, Phase I (Saipan) (here- 
after cited as NTLF Rpt Marianas, Phase I), Incl 
E, G-3 Rpt, p. 1. 



sonnel. On the Northern Troops and Land- 
ing Force Staff, the assistant chiefs of 
staff for both intelligence (G— 2) and sup- 
ply (G-4) were Army officers — Lt. Col. 
Thomas R. Yancey and Lt. Col. Joseph C. 

There were disadvantages to this cellu- 
lar fission, however unavoidable it may 
have been. First, there was a decided short- 
age of trained personnel, especially of spe- 
cial staff sections, officer assistants and 
trained clerks, draftsmen, and stenograph- 
ers; and second, a shortage of headquar- 
ters and corps troops already existed in V 
Amphibious Corps. 20 "In effect," as one 
commentator put it, "you have here an 
army and a corps trying to operate with 
a staff too small for a corps." 21 

All echelons prepared their plans simul- 
taneously, and the normal time sequence 
of planning from highest echelon down, 
with each subordinate basing his own plan 
on that of his immediate superior, was sel- 
dom achieved. For example, Headquarters 
Expeditionary Troops (Task Force 56) 
Operation Plan 3-44 was issued on 26 
April, whereas the next higher echelon, Ad- 
miral Turner's Headquarters Northern At- 
tack Force (Task Force 52) did not issue 
its plan until 21 May. 22 Again, it was not 
until 12 May that Admiral Spruance, who 
was superior to both Turner and Smith, 
came out with his operation plan for the 
Fifth Fleet. 23 

20 ibid., p. q. 

21 C&GS School, Ft. Leavenworth, Kans., Sec- 
ond Command Class, Recent Operations, Foragkr 
(Marianas), Annex C, p. 2. 

33 Northern Attack Force (TF 52) Attack Order 
A11-44, 21 May 44. As Commander, Expedition- 
ary Forces (TF 51), Turner issued his Operation 
Plan A 1 0-44 on 6 May 1944. 

2:1 Comdr Fifth Fleet Opn Plan Central 10-44, 
12 May 44; TF 56 Rpt Forager, Incl B, G-5 
Rpt, p. 5. 

Northern Troops and Landing Force 
Headquarters Operation Plan 3-44 of 1 
May summarized all previous plans from 
higher echelons and governed the tactical 
order of all troops in the proposed landings 
on Saipan. 24 The 4th Marine Division (re- 
inforced) was to land on Blue and Yellow 
Beaches, extending from the town of 
Charan Kanoa south almost to Agingan 
Point. Its first objective was to be a line 
inland from the beaches about 2,000 yards 
at the north and tapering down to the 
water's edge at the southern end. Then, 
on order, the division was to advance ra- 
pidly and seize A slito airfie ld and the sur- 
rounding terrain. \{Map I.)\ 

The 2d Marine Division was to land si- 
multaneously to the north of Charan Kanoa 
on Green and Red Beaches, seize the first 
commanding ground inland, and then ad- 
vance rapidly and capture Mount Tapot- 
chau and Mount 'Tipo Pale and the adja- 
cent ground. The XXIV Corps Artillery 
was to land on order on beaches to be 
designated and to execute missions as as- 

North of the 2d Marine Division's 
beaches, in the vicinity of Tanapag Har- 
bor, a naval force consisting of transport 
divisions carrying reserve regiments from 
the 2d and 4th Marine Divisions would 
conduct a diversionary demonstration to 
last from a half hour before sunrise to an 
hour after the main landing. 

Finally, the 1st Battalion, 2d Marines, 
was detached from its parent organization 
to perform a separate mission. Originally, 
the battalion was to land from destroyer 
transports (APD's) on Magicienne Bay on 
the southeast side of the island the night 
before the main landing on the west coast. 

24 NTLF Opn Plan 3-44 (Forager), 1 May 44. 



It would then move rapidly inland, at- 
tempt to seize Mount Tapotchau before 
daylight and hold on until relieved by the 
main elements of the 2d Marine Division. 
Later, on 7 May, this order was changed, 
and the battalion was to be prepared to 
land on Magicienne Bay or perhaps other 
beaches after the main landing had been 
effected and then move west and north to 
attack enemy positions from the rear. 25 
Eventually, the whole scheme was can- 
celed as impractical and involving excessive 

The final decision was in all probability 
the soundest one. To have committed a 
single battalion armed with nothing heav- 
ier than 60-mm, mortars against the for- 
midable defenses the Japanese had set up 
around Magicienne Bay would in all like- 
lihood have proved disastrous. As events 
turned out, it took the entire 2d Marine 
Division ten days to reach Mount Tapot- 
chau's summit. 26 

At headquarters of the 27th Division the 
problem of planning for landings on Saipan 
was seriously complicated because there 
was no certainty as to how the division 
would be employed. It was the corps re- 
serve and might be committed on Saipan 
only in part or piecemeal, might be re- 
served for later action on Tinian and 
Guam, or might not be used at all. In 
short, there were a large number and wide 
variety of possibilities, and operations of- 
ficers had to plan accordingly. Hence, 
Ralph Smith's G-3 (operations) section 

iri Ibid., Change 1, 7 May 44. 

- a These conclusions arc based on a concurrence 
of opinion of several planning and intelligence 
officers in the 2d Marine Division and the V 
Amphibious Corps. See Major Carl W. HofTman, 
USMC, Saipan: The Beginning of the End, His- 
torical Division, Headquarters, U.S. Marine Corps 
(Washington, 1950), p. 28. 

found it necessary to prepare a total of 
twenty-one complete plans for tactical em- 
ployment of the division. A few of these 
were discarded as higher headquarters 
made progress in outlining the details of 
their own plans. By the time the troops 
sailed from Hawaii it appeared that, if used 
on Saipan at all, the division would prob- 
ably be employed in one of three ways and, 
accordingly, three preferred plans were de- 
vised. The first contemplated a landing of 
two regiments (105th and 165th) on 
beaches at Magicienne Bay and a rapid ad- 
vance northwest across the island to cap- 
ture the seaplane base at Flores Point. The 
second envisaged a landing by the same 
two regiments on the beaches north of 
Charan Kanoa, on the left of the qd Ma- 
rine Division, followed by a northward 
thrust to Garapan Village. If either of 
these were executed, the third regiment 
( 1 06th Infantry) would act as floating re- 
serve. Plan number three called for the two 
assault regiments to go ashore at Tanapag 
Harbor and prepare to move southward to 
join forces with the 2d Marine Division. In 
this case, the 106th Regimental Combat 
Team (RCT) was to seize Maniagassa Is- 
land off Tanapag and support the main 
assault. 27 A final plan was made only after 
the division had sailed from Hawaii. On 
its arrival at Kwajalein the 106th Infantry 
was attached to the Southern Landing 
Force and ordered to prepare plans for a 
landing on Guam. 28 

In actual fact, all of the plans had to be 
abandoned early in the battle for Saipan. 
Although it cannot be said that all the 
laborious preparations by the 27th Divi- 
sion were entirely wasted, it is true that 

Hq 27th Inf Div Opn Plans I, II, III. 
Hq 106th Inf Forager Opn Rpt, p. 2. 



neither the division nor the corps head- 
quarters had on hand a detailed plan that 
exactly fitted the situation as it had de- 
veloped by the time the division was 

Preliminary naval and aerial bombard- 
ment of the Marianas was planned along 
lines by then well established in the Central 
Pacific theater. Landings at Tarawa and 
in the Marshalls left little doubt of the 
necessity for heavy preliminary pounding 
of the beaches from both the air and the 
sea if excessive American casualties were 
to be avoided. 2 " 

For Saipan, an impressive armada of 
ships and planes was allocated to do the 
job. A total of fifty-five ships was original- 
ly scheduled to deliver fire against the 
main island: 7 fast battleships from Marc 
Mitscher's fast carrier force, 4 old battle- 
ships, 2 heavy cruisers, 3 light cruisers, 15 
destroyers, and 24 LCI gunboats (LCI 
(G)'s). Simultaneously with this bombard- 
ment, the smaller island of Tinian was to 
be subjected to similar fire from an 
additional 33 ships, including 3 old battle- 
ships, 4 heavy cruisers, 2 light cruisers, 7 
destroyers, and 24 LCl(G)'s. 30 

Two days before the scheduled landing, 
fast battleships and destroyers of Task 
Force 58 were to bombard Saipan and 
Tinian, destroy aircraft, put airfields out 
of commission, destroy coast defense and 
antiaircraft batteries, burn ofl cane fields 
in the landing area, deliver antipersonnel 
fire, and, finally, cover mine-sweeping oper- 
ations off the western shore line. Next day, 
old battleships and smaller fire support 
ships of Turner's Task Force 52 were 
scheduled to deliver counterbattery fire, 

area bombardment, and interdiction fire, 
commencing at daybreak and continuing 
throughout the day, Their primary mission 
was to destroy as many coast defense guns, 
antiaircraft batteries, artillery weapons, 
and other enemy defenses and personnel 
as possible. Ships were directed to remain 
well beyond the range of enemy shore bat- 
teries on that day, which meant in effect 
that their fire would be delivered at ranges 
in excess of 10,000 yards. 31 They were 
instructed to pay particular attention to 
gun positions at Magicienne Bay and to 
the beach defenses and installations on the 
selected landing beaches on the west coast. 
Also, they were to cover mine-sweeping op- 
erations and beach reconnaissance by the 
underwater demolition teams, whose job 
it was to inspect the beaches and ap- 
proaches thereto for mines, underwater 
obstacles, and explosives. Simultaneously, 
ships of Admiral Conolly's Task Force 53 
were to work over neighboring Tinian in 
much the same manner, although these 
vessels were to conserve most (80 percent) 
of their ammunition allowance for preas- 
sault bombardment of Guam. 32 

For D Day (15 June) on Saipan the 
schedule of fires was to be stepped up 
sharply, with particular attention to be 
paid to the landing beaches. Counterbat- 
tery fire was to commence at dawn and to 
cover known and suspected positions of 
enemy coast defense guns and antiaircraft, 
dual-purpose, and field artillery batteries 
both on Saipan and on Tinian. Ships were 
to be in position to bombard beach de- 
fenses and possible flanking positions, with 
close-range fire to commence at the low- 

29 Sec Crowl and Love, Gilberts and Marshalls. 
30 NTLF Opn Plan 3-44 (Foraoer), i May 
44, Annex Charlie, pp. 1-2. 

31 TF 56 Rpt Forager, Rpt on Naval Gunfire 
Support, p. 10. 

32 Ibid., pp. 2-4; TF 52 Attack Order An-44, 
21 May 44, Annex G, pp. 3-9. 



water line and extend 400 yards inland. 
Area bombardment of secondary defenses 
such as supply installations, barracks, and 
bivouac areas was to be continued, as was 
supporting bombardment of Tinian by 
Task Force 53. Shortly before the sched- 
uled landing hour (H Hour) on Saipan, 
close supporting fires were to be delivered 
against the Charan Kanoa beaches and in 
the Tanapag Harbor area, the latter being 
in support of the demonstration landing 
and therefore on a smaller scale. All naval 
gunfire, except for counterbattery fire 
necessary to the protection of ships and 
landing craft, was to cease for a half hour 
(between H minus 90 and H minus 60) 
to permit a low-altitude aerial strike on the 
beaches, and then resume for the hour be- 
fore the landing. For the hour remaining 
before the troops were scheduled to touch 
shore, battleships, cruisers, and destroyers 
were directed to move in to close range and 
bombard the selected landing beaches and 
adjacent installations. 33 

Then, just before the scheduled landing 
hour, when the assault troops were as- 
sembled in the leading waves of amphibian 
tractors, twenty-four LCI gunboats, equip- 
ped with rockets and 20-ram, and 40-mm. 
guns, were to move slowly forward toward 
the beach in line-abreast formation, just 
ahead of the first wave of amphibian 
tanks. As the LCI's reached the line where 
the heavier fire support ships lay to, they 
were directed to open fire on the beach 
areas with their 40-mm. guns. Those off 
the northern beaches (Green and Red) 
were to stop dead in the water at this line, 
let the leading waves pass through them, 
and continue to fire as long as safety to 
the landing craft permitted. No rockets 

were to be fired by the northern group of 
gun boats since the reef in this area would 
keep them out of effective range of the 
beach (1,100 yards). At the southern 
beaches (Blue and Yellow), LCI(G)'s 
were ordered to proceed at a distance two 
hundred yards ahead of the first landing 
craft until they reached a line 1,000 yards 
off the beach, then fire their rockets and 
40-mm. guns as long as safety allowed. 34 
Just as impressive as the plans for pre- 
paratory naval fire were those for preland- 
ing aerial bombardment. Mitscher's Task 
Force 58 had made its first strike against 
the islands of Guam, Rota, Tinian, and 
Saipan on 23 February , 3S and thereafter 
on the occasions when aerial reconnaissance 
missions were flown across the islands 56 
some bombs were released, although with 
dubious results. According to the original 
plans, however, not until two days before 
the scheduled landing on Saipan would a 
heavy and prolonged aerial bombardment 
of that island and Tinian be undertaken. 
The fast carriers of Mitscher's force, work- 
ing in conjunction with escort carriers un- 
der command of Admiral Turner, would 
undertake this task, 37 One D minus 2 (13 
June) planes from the fast carrier force 
were to make fighter sweeps on airfields on 
both Saipan and Tinian to destroy enemy 
aircraft. On the same day thirty-three 
planes would deliver counterbattery fire 
against guns firing on the mine sweepers. 
Combat air patrol and antisubmarine 
patrol missions were to be flown simultan- 
eously. The next day a more intensive pro- 

x:i CTF 52 Attack Order A11-44, App. 6A to 
Annex C. 

34 Ibid., App. 6A to Annex C, p, 2; App. 5 to 
Annex J, p. 14, 

3n CINCPAC-CINCPOA Opus in POA, Feb 44. 

S8 See below, pp. 50-51. 

37 GTF 5 1 Opri Plan A 10-44, Annex F; CTF 
56 Opn Plan 3-44, Incl A to TF 56 Rpt Foragkr, 
Annex D; GTF 56. i Opn Plan 3-44, Annex D. 



gram of destruction was to be undertaken. 
Inland coast defense and dual-purpose and 
antiaircraft guns were to be bombed 
heavily. Cane fields not already burned 
were to be fired. Other priority targets 
were inland defense installations and struc- 
tures, the buildings around Aslito airfield, 
and communications and transportation 
facilities on the west coast of Saipan in- 
cluding small craft, radio stations, observa- 
tion towers, railroad and road junctions, 
and vehicles. The same day six smoke 
planes were to provide protection for un- 
derwater demolition teams operating close 
offshore, if necessary. Also, vertical photo- 
graphs were to be made of all beaches from 
Tanapag Harbor to Agingan Point. 

On 15 June, in addition to continuing 
most of the above duties, a heavy half-hour 
aerial attack on both islands was to be 
carried out and to be terminated only one 
hour before the scheduled landings. Dur- 
ing this period naval gunfire was to be 
lifted so that planes could fly in low for 
precision bombing and rocketing of enemy 
installations. A total of 60 fighters, 51 dive 
bombers, and 54 torpedo bombers were to 
take part in this final preliminary satura- 
tion attack. Thereafter, until the carriers 
were withdrawn, the carrier-based planes 
would act as aerial observers for land-based 
artillery, make photographic sorties, lay 
smoke on request, and provide deep and 
close support for the troops ashore. 

For the critical movement of assault 
troops from ship to shore the plans fol- 
lowed, with some variation, the pattern 
used so successfully in the Marshalls oper- 
ation. 38 Astern of the LCI gunboats, 
amphibian tanks (LVT(A)'s) would con- 

stitute the bulk of the leading wave. They 
mounted either 75-mm. howitzers or 37- 
mm. guns plus machine guns, and their 
first job was to lead the waves of assault 
amphibian tractors ( LVT's ) from the reef's 
edge to the shore line. The LVT(A)'s 
would provide the only close fire support 
for the assault troops during the critical 
few minutes between the time that naval 
gunfire and aerial bombardment were com- 
pelled to lift and the time that the 
infantrymen actually hit the beach's edge. 
Moreover, for the Saipan landing the mis- 
sion of the amphibian tanks was not to 
cease at the shore line. On the 4th Marine 
Division beaches, the tanks of the 708th 
Amphibian Tank Battalion (Army) were 
to push inland approximately 1,500 yards 
to the first objective line and set up a 
perimeter defense closely supported by in- 
fantry in the amphibian tractors. 39 To the 
northward, in the 2d Marine Division's 
zone of action, the 2d Armored Amphibian 
Battalion (Marine) was ordered to move 
four companies of its amphibian tanks in- 
land only about three hundred yards to 
the tractor control line and there cover the 
debarkation of assault troops from its 
LVT(A)'s. Thereafter, most of the am- 
phibian tanks were to remain under cover 
and engage targets as far inland as 1,500 
yards, but only on call from the infantry. 40 
Thus, LVT(A)'s were scheduled to pro- 
ceed beyond the beaches' edge and to act, 
to all intents and purposes, as land tanks 
until such time as heavier tanks could be 
brought ashore. This was an innovation in 
amphibious techniques and one that, as 

:iS See Crowl and Love, Gilberts and Marshalls, 
pp. 172-82. 

"» NTLF Rpt Marianas, Phase I, Incl H, LVT 
Rpt, Rpt of 708th Armd Amph Tank Bn, 5 Jul 

44, P- 3- 

40 Ibid., Incl G, Rpt of 2d Armd Amph Bn, r 
Jul 44, p. 1. 



events developed, proved to be of dubious 
merit. 41 

Training and Rehearsals 

With the conclusion of the Marshalls op- 
eration, it became apparent that the future 
promised a shift in the Central Pacific Area 
from atoll warfare to operations on larger 
land areas that were both mountainous 
and jungle-covered. Hence, even before the 
official warning orders came down from 
corps headquarters, all three divisions as- 
signed to Northern Troops and Landing 
Force had commenced training their troops 
to meet the particular conditions that the 
forthcoming campaign would impose. 

The 27th Division, stationed on Oahu, 
made an early study of the methods of 
burning sugar cane and the movement of 
foot troops through freshly burned fields. 
Groups from all of the infantry regiments 
conducted exercises in methods of burning 
fields, cutting passage through them, and 
the movement of large and small numbers 
of troops through standing and freshly 
burned cane. 

The division also concentrated heavily on 
training its men in combined tank-infantry 
operations. All infantry companies engaged 
in field exercises involving the use of tanks 
in direct support, a particularly important 
exercise for the Army division since its tank 
battalions were not organic but were spe- 
cifically attached to the infantry for the 
Marianas operation. Other specialized 
training included intense education in am- 
phibious communications procedures. The 
295th Joint Assault Signal Company 
(J AS CO) was attached to the division suf- 
ficiently far in advance to allow for 
thorough familiarization of the infantry 

battalions with the functions and abilities 
of the various JASCO teams. 

All units were instructed in the proper 
organization and plan of fires for a night 
perimeter defense. A period of five weeks 
was devoted to the study and practice of 
methods of loading 105-mm. howitzers in 
amphibian trucks (DUKW's). Combat en- 
gineers were instructed in the use of flame 
throwers and demolitions for the reduction 
of fortified positions. The 27th Cavalry Re- 
connaissance Troop conducted rubber boat 
training with emphasis on beach reconnais- 
sance, hydrographic studies, and night 
landings. 42 

The 2d Marine Division boasted excel- 
lent training facilities in the vicinity of its 
"Camp Tarawa" on the main island of 
Hawaii. It too held special exercises in 
the techniques of fighting through sugar 
cane. Also, the jungle and mountainous 
terrain on Hawaii approximated the type 
that the division would meet on Saipan and 
was ideal for the simulation of realistic 
combat conditions. 43 

The 4th Marine Division fared less well. 
Its camp site on Maui was new, its living 
and training facilities were incomplete. 
Hence, camp construction and training 
had to be carried out simultaneously — a 
situation that, though common enough in 
the Pacific, was never desirable. Neverthe- 
less, by instituting emergency measures for 
the acquisition of suitable land and 
through co-operation with the Navy and 
Army authorities, "a fairly satisfactory 
schedule of individual, unit and combined 
training was completed," according to the 
division's commanding officer. 44 

See below, pp. 85-87. 

42 27th Inf Div Rpt Forager, G-3 Combat Rpt, 

PP- '-5- 

4S 2d Marine Div SAR, Phase I, Forager, p. 1. 
44 4th Marine Div Opns Rpt Saipan, p. 6. 



The chief training problem facing XXIV 
Corps Artillery was that of converting two 
coast artillery battalions into field artillery 
battalions. The XXIV Corps Artillery was 
not activated as a unit until the end of 
March 1944. At that time the 225th Field 
Artillery Group was alerted to take part in 
the Marianas operation and relieved of its 
defensive mission on Oahu. Two of its field 
artillery battalions were detached, and the 
32 d Coast Artillery Battalion and 2d Bat- 
talion, 55th Coast Artillery, were attached. 
As finally organized, the XXIV Corps Ar- 
tillery was made up of one 155-mm. 
howitzer group of two battalions and one 
155-mm. gun group of two battalions 
drawn from available field artillery person- 
nel and supplemented by coast artillery 
personnel. Because of its infancy as an or- 
ganization and the lack of combat experi- 
ence of most of its personnel, adequate 
training for the unit was urgent. 

Coast artillery officers were immediately 
given an intensive education in basic field 
artillery methods, and a similar program 
for enlisted personnel followed. Demonstra- 
tions held by field artillery batteries and 
battalions were followed by four field exer- 
cises per week. On 1 May two batteries of 
the 53d Field Artillery Battalion were 
loaded on an LST and taken to Maui to 
experiment with methods of loading 155- 
mm. guns and to obtain training in unload- 
ing across sandy beaches. Experiments 
were also conducted in loading the 155's on 
the smaller LCT's (landing craft, tank). 

Because of the shortness of time and its 
relative lack of basic training, XXIV Corps 
Artillery did not participate in the final 
grand rehearsal. Instead, the two-month 
intensive training period culminated in a 

corps artillery field exercise held during the 
rehearsal period. 45 

On 14 May ships carrying the two Ma- 
rine divisions with their full loads of 
equipment rendezvoused in the area of 
Maalea Bay, Hawaii, for final rehearsals 
before shoving off to Saipan. LVT's and 
other amphibious craft were launched; the 
assault battalions practiced ship-to-shore 
movements; shore party team personnel 
and beach parties were landed with their 
communications equipment; artillery was 
beached and dragged ashore. On 1 6 and 1 7 
May each of the divisions made a co- 
ordinated landing on the island of Maui 
and battle conditions were simulated as far 
as was practicable. However, in view of 
the fact that the island was populated, ship 
and aerial bombardment had to be "con- 
structive" only. Moreover, the landing 
beaches were separated and maneuver area 
ashore was extremely limited, preventing 
rehearsal of co-ordinated movements inland 
and any extensive deployment of troops 
once they had reached the shore line. As 
an exercise in ship-to-shore movement the 
rehearsal was useful, but it failed to give 
the troops an adequate foretaste of the 
problems involved in consolidating a beach- 
head once they had landed. 46 

Finally, on 19 May, a simulated landing 
was made jointly by the two Marine di- 
visions on the nearby island of Kahoolawe. 
This time troops approached the shore un- 
der actual cover of naval and aerial fire. 
On reaching a line 300 yards from the 
beaches they turned back, but in every 
other respect the exercise was a full-dress 

45 XXIV Corps Arty Final Rpt on Forager 
Opn, Phases I and III, S-3 Rpt, pp. 1-3. 

40 2d Marine Div SAR, Phase I, Forager, Sec. 
Ill, pp. 1-3; 4th Marine Div Opns Rpt Saipan, 
Incl A, pp. 7-8. 



rehearsal of the plans for the forthcoming 
landing on Saipan, with units, positions, 
intervals, distances, and other details as 
prescribed in the operation plans. This was 
followed immediately by a second exercise 
in which air and naval gunfire did not take 
part. Then troops were re-embarked in the 
vessels in which they were scheduled to sail 
overseas and returned to their respective 
rehabilitation areas. 47 

The only incident that had marred last- 
minute training was the loss of three deck- 
loaded LCT's over the sides of the LST's 
that were carrying them to the rehearsals. 
This resulted in twenty-nine casualties to 
the 2d Marine Division. Moreover, two of 
these LCT's had been specially equipped 
with 4.2-inch mortars. Plans called for their 
employment on Saipan as support ships to 
supplement the rocket -firing LCI's in the 
last minutes between the lifting of heavy 
ships' fire and the landing of troops. Their 
loss during rehearsals prevented the equip- 
ment from being tested until a later opera- 
tion and deprived the assault troops of that 
much additional naval support. 41 * 

From 18 to 24 May the 27th Division 
(minus its artillery), fully loaded on three 
transport divisions, conducted similar re- 
hearsals. The exercise emphasized the tech- 
nique of debarking and landing a large 
number of troops with a limited number 
of boats, a situation thought likely to occur 
if reserve troops had to be landed at all on 
Saipan. Ship-to-shore communication was 
established although, as in the case of the 
two Marine divisions, no supplies were un- 

loaded since all ships had already been as- 
sault loaded for the actual landing, 4 '"* 

Loading and Embarkation 

The task of carrying three reinforced di- 
visions and almost seven thousand corps 
and garrison troops with all their supplies 
and equipment over a distance of 3,200 
miles from Hawaii to Saipan was the heav- 
iest yet imposed upon the Navy in the 
Pacific war. To accomplish it, Admiral 
Nimitz assembled a flotilla of no naval 
transport vessels of all varieties — 37 troop 
transports (APA's and AP's), 11 cargo 
ships (AKA's and AK's), 5 LSD's (land- 
ing ships, dock), 47 LST's, and 10 

In addition, a whole division of Liberty 
ships had to be organized to transport the 
1 06th Regimental Combat Team because 
of the scarcity of Navy troop transports in 
the area. 01 

All together, a total, of 74,986.6 measure- 
ment tons of cargo representing 7,845,194 
cubic feet was loaded. By comparison, dur- 
ing the invasion of Kwajalein in January 
1944, only 49,283 tons were carried in the 
assault shipping. 52 Nimitz' operation plan 
provided that assault and garrison forces 
should be allowed 32 days of Class I sup- 
plies (rations), 20 days of Class II (organ- 
izational and individual equipment), 20 
days of Class III (fuels and lubricants), 
20 days of Class IV (miscellaneous), and 

17 Ibid. 

* H 2d Marine Div SAR, Phase I, Forager, Sec. 
Ill, p. 1 ; United States Army Forces Central Pa- 
cific Area (USAFICPA), Participation in the 
Marianas Operation, Jun-Sep 44, Vol. I, p. 58. 

49 37th Inf Div Rpt Forager, G-3 Combat Rpt, 

P- 5- 

n " Commander in Chief, U.S. Fleet, Amphibious 
Operations, Invasion of the Marianas, June to 
August 1944 (Washington, 1944) (COMINCH 
P-007), V-i. 

01 TF 56 Rpt Forager, Incl G, TQM Rpt, p. 7. 

S2 COMINCH P-007, V-4; TF 56 Rpt For- 
ager., Incl G, TQM Rpt, p. 3. 



7 (CINGPOA) units of fire for ground 
weapons and 10 for antiaircraft weapons. 53 

Staging areas for the three divisions were 
widely separated and not all were conven- 
iently located. Ships assigned to the 2d 
Marine Division loaded at Hilo on the main 
island of Hawaii, those of the 4th Marine 
Division at Kahului, Maui, and those of 
the 27th Division at Oahu. The ports of 
Kahului and Hilo were ill suited to loading 
the two Marine divisions efficiently. The 
piers at each could berth only four ships 
alongside the dock at one time. There were 
not enough dock cranes, stevedore equip- 
ment, and warehouses. There were no dock 
lighting facilities, and it was difficult for 
LST's to beach properly. 04 Also, Hilo was 
some sixty miles away from the 2d Marine 
Division's Gamp Tarawa, which compli- 
cated the problem of loading both troops 
and equipment. 55 

Standard combat unit loading proce- 
dures were followed as a rule, but shipping 
shortages sometimes made this impos- 
sible/' 8 This was especially true of supplies 
and equipment belonging to V Amphibious 
Corps troops, XXIV Corps Artillery, and 
garrison troops. The last available AP in 
the Pacific (USS G. F. Elliott), two AK's 

S; < CINCPAC-CINCPOA Opn Plan 3-44, Annex 
C ; TF 56 Rpt Forager, Annex E, G-4 Rpt, pp. 
4—5 ; for CINCPOA units of fire, see Appendix B, 

84 TF 56 Rpt Forager, Annex G, TQM Rpt, 
PP- 13-16. 

55 2d Marine Div SAR, Phase I, Forager, Sec. 
II, p. ,. 

5,! Combat unit loading is defined as the method 
"in which certain units, selected because of their 
probable employment to meet tactical situations 
immediately upon landing, are completely loaded 
in a single transport with at least their essential 
combat equipment, transportation, and supplies 
available for immediate debarkation with the 
troops." (Office of Naval Operations, Division of 
Fleet Training, Landing Operations Doctrine 
United States Navy, 1938 (FTP 167), p. an.) 

(USS Hercules and USS Jupiter), and two 
LST's were assigned to lift these units, but 
the shipping space was inadequate. Excess 
personnel (approximately 4,000) were dis- 
tributed among the transports carrying the 
two Marine divisions. As a result most of 
these units, especially the XXIV Corps 
Artillery, were separated from their cargo. 
In other words these particular units were 
"convoy unit loaded" — which was highly 
undesirable from the point of view of tacti- 
cal disposition. 57 

Even after parceling out more than half 
of its attached troops to the ships carrying 
the Marine divisions, V Amphibious Corps 
still did not have enough room aboard its 
own vessels. It was impossible to combat 
load its cargo. To have tried to vertically 
load each of the twenty-five units carry- 
ing cargo on the corps ships would have 
meant leaving from 25 to 35 percent of the 
cargo behind. The upshot was that a top 
priority was assigned to corps artillery 
and the corps signal battalion, and the re- 
mainder of the units' equipment was 
stowed wherever it could be fitted in. B8 

As early as i May 1 944, Holland Smith's 
headquarters had ordered 25 to 50 percent 
of all supplies and two to five units of fire 
to be palletized. 59 The object was to per- 

This means, in effect, that materials most urgently 
needed will be stowed near the top of ships' holds 
and immediately under hatch covers. Lower prior- 
ity supplies will be stowed lower in the holds and 
away from the hatch covers where they cannot be 
unloaded without first removing everything be- 

nT Convoy unit loading is defined as the method 
in which "the troops with their equipment and 
supplies are loaded in transports of the same con- 
voy, but not necessarily in the same vessel." FTP 
167, p. 211. 

68 TF 56 Rpt Forager, Incl G, TQM Rpt, pp. 

59 V Phib Corps Order 3-44, 1 May 44, par. 
(g). Pallets are wooden, sledlike structures to 
which supplies can be lashed. 



mit rapid transfer from the beach to inland 
dumps by using tractors to drag the pal- 
lets instead of resorting to the older method 
of trucking loose supplies inland employ- 
ing manpower to load the trucks. From 
the outset, the two Marine divisions were 
lukewarm toward the project, and in the 
end loaded only a few pallets. The 27th 
Infantry Division, however, had had sever- 
al months' experience in handling palletized 
cargo and was enthusiastic about this 
technique for loading and unloading sup- 
plies. 00 The division not only complied 
with corps orders, but went beyond it and 
palletized between 80 and 90 percent of 
all supplies. 61 

One reason for the Marine divisions' 
failure to follow suit was inexperienced 
labor and a shortage of equipment. The 4th 
Marine Division reported that it could 
procure only enough material to palletize 
10 to 15 percent of all supplies and that 
the job was done so poorly that some pal- 
lets broke down during handling. 62 In the 
end, the division decided that palletization 
of supplies at least for the initial stages of 
the assault was not worth the trouble. The 
marines argued that palletized supplies 
took up too much space aboard ship, were 
difficult to transfer from one type of land- 
ing craft to another, and required too 
much extra equipment. Furthermore, it 
was contended that pallets were not prac- 
tical where dumps were located more than 
500 yards inland and where reefs were 
encountered. 03 

Neither corps headquarters nor the 27 th 
Division agreed. Holland Smith's transport 
quartermaster maintained that the "rea- 
sons for palletization overbalance the neg- 
ative effects," and cited as the primary 
benefits the rapid unloading of landing 
craft at the beaches and the release of large 
working parties formerly engaged in trans- 
ferring cargo from landing craft to 
trucks. 64 The 27th Division headquarters 
was so enthusiastic about the process that 
it diverted from training one and sometimes 
two companies of infantry in addition to a 
platoon of engineers for a period of six 
weeks just to palletize supplies. 85 

Amphibian tanks and tractors, the all- 
important vehicles of assault, were as usual 
transported aboard LST's. Each LST car- 
ried seventeen LVT's, loaded in two rows of 
eight with the odd one secured on the 
ramp. By loading LVT's in this manner, 
about fifteen feet of clear space remained 
on the after portion of the LST tank deck, 
and emergency supplies were "preloaded" 
thereon. In addition to the amphibian ve- 
hicles, each LST carried more than 300 
marines from Hawaii to Eniwetok. There, 
they received fifty to seventy-five more 
from transports to fill the complement of 
the assault waves. A serious LST shortage 
almost occurred when six were destroyed 
by fire at Pearl Harbor on 21 May. How- 
ever, LST's originally assigned to the gar- 
rison force were used as substitutes, and 
loading and embarking was only delayed 
twenty-four hours. 66 

One impediment to well-planned and 
well-co-ordinated combat loading was that 

00 TF 56 Rpt Forager, Incl G, TQM Rpt, p. 


61 27th Inf Div Rpt Forager, G-4 Rpt, Supply 
Phase of Forager Opn, Incl 3, p. 1. 

02 4th Marine Div Opns Rpt Saipan, Annex D, 
Supply and Evacuation, p. 6, 

63 Ibid,, Annex E, p. 33. 

04 TF 56 Rpt Forager, Annex G, TQM Rpt, 

P- 33- 

65 27th Inf Div Rpt Forager, G-4 Rptj Supply 
Phase of Forager Opn, p. 3. 

«" TF 56 Rpt Forager, Incl G, TQM Rpt, pp. 

TO, 17. 



troop transport quartermasters too often 
received insufficient or inaccurate informa- 
tion on the characteristics of the ships 
assigned to them. Precise data on the 
location, size, and shape of ships' holds, 
the number and location of hatches and 
winches and other equipment, plus myriad 
other details concerning ship structure are 
essential to proper combat loading. This 
was not always forthcoming. New ships ar- 
rived at the very last moment, and there 
was little or no time available to obtain 
correct ships' characteristics. 67 For two 
AP's (USS Storm King and USS John 
Land) assigned to the 4th Marine Division, 
no characteristics were obtainable before 
atmal loading. The division's supply sec- 
tion had been instructed to assume that 
these vessels' characteristics were similar to 
those of another AP, USS LaSalle. Upon 
arrival of the ships, it was discovered there 
was no such resemblance, that the new 
ships were not entirely suitable for combat 
loading, and that the winchmen were in- 
experienced and too few in number to cope 
with the problems at hand. Hence, many 
valuable items of equipment, especially 
twenty-five 2j/a-ton cargo trucks, had to be 
left behind. 68 

In spite of these and kindred difficulties, 
the three divisions met Admiral Turner's 
loading schedule. By 14 May both the Ma- 
rine divisions were aboard their transports 
and ready to depart for rehearsals, com- 
pletely loaded except for a few last-minute 
items. By 18 May the 27th Division was 
also set to go. 68 

After a brief period of rehabilitation 
following rehearsals, all units of Northern 
Troops and Landing Force once again 
boarded their ships and prepared to set sail 
for the final ordeal. The slower LST's car- 
rying assault elements of the two Marine 
divisions sortied from Pearl Harbor on 25 
May. On 29 and 30 May two groups of 
naval transports followed. All ships carry- 
ing the assault troops rendezvoused at 
Eniwetok, where last-minute intelligence 
data was disseminated and additional 
troops assigned to the initial landing waves 
were transferred from transports to LST's. 
By 1 1 June the last of the attack transports 
had weighed anchor in Eniwetok lagoon 
and the mighty convoy, split into four 
separate groups, was steaming westward 
through hostile waters toward still more 
hostile shores.* 70 Well to the rear came 
the transport and tractor (LST) groups 
carrying the reserve troops, the 27th In- 
fantry Division. These had sailed from 
Pearl Harbor between 25 May and 1 June 
and had rendezvoused at Kwajalein, There, 
the 106th Regimental Combat Team was 
informed that it would undoubtedly be de- 
tached to the Southern Attack Force for 
the invasion of Guam. Otherwise, the voy- 
age for all units was uneventful. 71 

The Prospects Ahead: 
Intelligence of the Enemy 

While still at anchor in Eniwetok, the 
intelligence section of Headquarters, Ex- 
peditionary Troops (Task Force 56), re- 
ceived a final batch of aerial photographs 

«" Ibid., Incl G, TQM Rpt, p. 17. 

68 4th Marine Div Opns Rpt Saipan, Annex D, 

P- 3- 

69 TF 56 Rpt Forager, Incl G, TQM Rpt, p. 
16, and Incl JJ; 27th Inf Div Rpt Forager, G-3 
Rpt, p. 5. 

70 2d Marine Div SAR, Phase I, Foraqer, Sec. 
IV, pp. i-a; 4th Marine Div Opns Rpt Saipan, 
p. 12. 

71 27th Inf Div Rpt Forager, G-3 Rpt, p. 6; 
27th Inf Div Arty Rpt Forager, p. 6. 



of Saipan and the southern Marianas. 
These had been made on 28 May for Sai- 
pan and on 29 May and 7 June for Guam. 
They were disseminated to the two Marine 
divisions before their departure from Eni- 
wctok on 1 1 June, although the initial 
assault elements aboard the LST's had left 
before the new information could reach 
them. Hence, the leading waves of troops 
would make their landings on the basis of 
information of the enemy situation as de- 
rived from photographic sorties flown on 1 8 
April. 72 

A final G-2 "Summary of the Enemy 
Situation" was prepared by Holland 
Smith's intelligence section on 1 3 June and 
represents the last-minute estimate of ene- 
my potentialities in the Marianas before the 
actual landing. 73 This document predicted 
that the Japanese had on Saipan alone 
from 15,000 to 17,600 troops, with an 
additional 10,150 to 10,750 on nearby Tjn- 
ian. Of the total, 9,100 to 11,000 were 
thought to be ground combat troops located 
on Saipan. The rest of the garrison, it was 
believed, was made up of air base person- 
nel, maintenance and construction person- 
nel (including Koreans), and a home 
guard. This represented a considerable 
increase over an estimate made a month 
earlier (9 May), which put the total num- 
ber of enemy troops on Saipan at 9,000 to 
10,000 and on Tinian at 7,500 to 8,500. 74 

Saipan had three airfields in varying 
stages of preparedness. Aslito Naval Air 
Station in the south was 3,600 feet in 
length and believed to be fully operational; 
an emergency landing strip 3,280 feet in 
length had been sited in the area of Charan 

73 TF 56 Rpt Forager, Incl D, G-2 Rpt, p. a. 

7;l Ibid., Annex F, pp. j— 1 5, 

71 NTLF Rpt Marianas, Phase I, G-a Rpt, App. 
A, G-a V Phib Corps Special Study of Enemy 
Strength in the Southern Marianas, 9 May 44. 

Kanoa; and at Marpi Point a large airfield 
(4,300 feet) was still under construction 
and was considered to be nonoperational. 
In addition, a major seaplane base at 
Flores Point in Tanapag Harbor was 
thought to be fully operational. 

The latest estimate of air strength on Sai- 
pan before the carriers' strikes was a total 
of 152 aircraft. However, on n and 12 
June 140 aircraft were destroyed on Sai- 
pan, Tinian, and Guam, and since no aerial 
opposition was encountered at Saipan on 
13 June, Japanese aerial resistance from 
Saipan was thought unlikely. 75 

The newest photographs of Saipan re- 
vealed several significant increases in the 
number of gun installations since 18 April, 
when the last photographic sortie had been 
flown. The most notable of these were an 
increase of 32 percent in the number of 
heavy antiaircraft guns, 28 percent in me- 
dium antiaircraft guns, and 37 percent in 
machine guns. 78 

The following tabic indicates the number and 
type of enemy installations estimated to be em- 
placed on Saipan as of 29 May 1944: 

Blockhouses . ..i 

Possible blockhouses 1 

Coast defense guns — — -«3 

Possible coast defense guns a 

Dual-mount dual-purpose guns 5 

Single-mount dual-purpose guns .....n 

Single-mount heavy AA 49 

Possible single-mount heavy AA 9 

Single-mount heavy AA emplacements 5 

Single-mount medium AA 134 

Possible single-mount medium AA 8 

Single-mount medium AA emplacements .1 

Covered artillery emplacements 3 

Pillboxes 37 

Possible pillboxes ... 4 

Machine guns (20-mm. or under) AA 264 

Possible machine guns (20-mm. or 

under) AA 6 

Machine gun emplacements 4 

Empty emplacements .. 23 

75 TF 56 Rpt Forager, Incl D, G-2 Rpt, Annex 
F, p. 1. 

7<! Ibid., pp. 5-6. 



Radar a 

Possible radar i 

Searchlights 31 

Unidentified installations 4 

According to intelligence estimates, the 
preferred landing beaches off Charan 
Kanoa were defended by a well-developed 
system of trenches, tank traps, pillboxes, 
and machine guns. It was assumed that in- 
fantry elements on the island would be as- 
signed chiefly to the defense of this area. 
The absence of extensive field fortifications 
and the presence of heavy-caliber weapons 
between Garapan and Flores Point sug- 
gested to intelligence officers that the de- 
fense of that area would be primarily the 
responsibility of artillery and antiaircraft 
elements of whatever guard forces, special 
landing forces, and antiaircraft units that 
were stationed on the island. Intelligence 
officers also believed that the machine guns 
around Aslito field, at the southern end of 
Charan Kanoa strip, and on the eastern 
end of Marpi Point would probably be 

manned by similar elements and by air base 
defense antiaircraft personnel. 

Intelligence also led the officers to believe 
that the enemy probably had a tank de- 
tachment or at least an amphibious tank 
unit on Saipan. This, plus other factors, 
suggested that the Japanese contemplated 
a strong defense at the shore line combined 
with a mobile defense in the area behind 
the preferred landing beaches. 77 

The last assumption was essentially cor- 
rect, even if some of the detailed estimates 
as to the number of enemy troops and in- 
stallations proved to be well under the 
mark. At any rate, nothing in the last- 
minute intelligence surveys indicated that 
a basic change in the preferred landing 
plans was necessary. The die was cast. Un- 
der mild skies and through gently rolling 
seas the advance groups of troop-laden 
ships moved in slow procession toward the 

77 Ibid., p. 7. 


The Enemy 

Prewar Japanese Activities 
in the Marianas 

At the termination of World War I Ja- 
pan, as one of the Allied powers, was 
awarded a Class C mandate over all of the 
islands and atolls north of the equator that 
had formerly been in the possession of the 
German Empire. These included the Mar- 
shalb, the Carolines, the Palaus, and the 
Marianas except for Guam. Under the 
terms of Article 22 of the Covenant of the 
League of Nations, Japan agreed to refrain 
from "the establishment of fortifications or 
military and naval bases" in her newly ac- 
quired territories. Eleven years later Japan 
gave the required two years notice of her 
intention to withdraw from the League and 
did so officially on 27 March 1935. 1 

Whether or not the Japanese made any 
active effort to fortify or garrison the man- 
dates before 1933 remains in doubt, al- 
though their policy of excluding foreign 
visitors from these scattered islands inevit- 
ably raised suspicions in the minds of 
interested westerners as to what was going 

1 Denys P. Myers, Handbook of the League of 
Nations (Boston and New York, World Peace 
Foundation, 1935), p. 378; International Military 
Tribunal for the Far East (IMTFE), Interna- 
tional Prosecution Section (IPS) Doc. 6527, p. 13. 
The proceedings of the IMTFE and attached 
documents are filed in the Law Library, Office of 
the Judge Advocate General, Department of the 

on behind the silken curtain. In any event 
after her withdrawal from the League and 
before the outbreak of hostilities it is certain 
that Japan embarked on a program of mil- 
itary construction in the area. This was 
done in spite of the fact that Japan's seces- 
sion from the League was a unilateral act 
and in international law did not relieve her 
of accepted obligations under the League 
Covenant. 2 

From 1934 to 15 November 1940, the 
Japanese Government is known to have 
appropriated at least 14,456,800 yen 
($3,939,478) for construction in the Mari- 
anas alone. 3 For the year beginning 1 5 No- 

2 IMTFE Proceedings, pp. 39, 43, 205-16, 408- 
15; Earl S. Pomeroy, Pacific Outpost: American 
Strategy in Guam and Micronesia (Stanford, 
Calif., Stanford University Press, 1951). 

It should be noted that Japan was under no 
obligation to the United States not to fortify the 
mandated islands. The United States was not a 
signatory to the League Covenant. By the Five 
Power Treaty signed at Washington in 1922, 
Japan agreed with the United States and Great 
Britain to maintain the status quo in regard to 
fortifications and naval bases in certain of her 
island possessions, but the Pacific mandates were 
not included in the agreement. See Harold and 
Margaret Sprout, Toward a New Order of Sea 
Power (Princeton, N.J., Princeton University Press, 
1940), Chs. X, XIII, and App. B. 

3 For this chapter, dollar value of the yen has 
been obtained by averaging the dollar value of the 
yen for the years 1934-40. Figures for all con- 
versions from yen to dollars have been taken from 
Japan Statistical Yearbook; ig4g (Nihon Statisti- 
cal Association, 1949), p. 612. 



vember 1940, a minimum of 121,189,666 
yen ($28,406,858) was appropriated for 
construction in the Marianas, Carolines, 
and Marshalls, of which 15,605,885 yen 
($3,658,019) was allocated to the Mari- 
anas. Of the total of over thirty million 
yen ($7,032,000) spent in the Marianas, 
about 60 percent was assigned to Saipan, 
25 percent to Tinian, and 15 percent to 
Pagan. About 40 percent of the Saipan 
appropriation was for air installations, the 
remainder being allocated to fortifications, 
barracks, storage buildings, offices, water 
supply facilities, ammunition storage facil- 
ities, and communications stations. Practi- 
cally all of the money allocated to Tinian 
and Pagan was earmarked for airfield 
construction. 4 

In 1934 work began on Aslito airfield, 
located near the southern end of Saipan. 
Aslito was the principal Japanese air base 
in the Marianas, and its capture and de- 
velopment was to be the main objective of 
the American forces that invaded the 
Marianas in 1944. In addition to Aslito, a 
seaplane base in Tanapag Harbor, Saipan, 
was completed in 1935. Five years later 
nearby Tinian could boast an airfield cost- 
ing about 7.5 million yen ($1,758,000). 

Although various Japanese spokesmen 
after the close of World War II pretended 
that these airfields and other building ac- 
tivities in the mandates were undertaken for 
peaceful purposes, the pretense was a flimsy 
one. A close examination of appropriations 
made by the Japanese Government for con- 
struction in the Marianas in 1940 and 1941 

4 Unless otherwise noted, the sources of informa- 
tion on Japan's prewar fortifications in the man- 
dates are: Special Forces, Early Scries, Vols. 9 
and 10, in U.S. National Archives, World War II 
Seized Enemy Records, Record Group 242, NA 
12226 and NA 12255, WDC 160867 and WDC 

clearly indicates that Japan had launched 
an active program of military fortification 
of the mandates well before the actual out- 
break of hostilities. In November of 1940 
a sizable appropriation was made for the 
construction of "lighthouses" throughout 
the mandated islands. Each "lighthouse" 
came equipped with barracks, ammunition 
storage facilities, a command post, and a 
lookout station. Actually, of course, these 
were naval lookout stations. 

On Saipan, construction was fairly ex- 
tensive. Twelve "lighthouses" were con- 
structed at a cost of 1,333,333 yen 
($312,533). In February 1941, 100,000 
yen ($23,440) was set aside to build four 
gun positions of reinforced concrete, to be 
completed by the end of July. During 1941 
almost 700,000 yen ($164,080) was de- 
voted to the construction of the Saipan 
branch of the 4th Fleet Naval Stores De- 
partment, including ammunition storage 
sheds with a floor area of 800 square 
meters. Also during 1941, almost 800,000 
yen ($187,520) was earmarked for con- 
struction of communications facilities, in- 
cluding receiving and sending stations, 
radio direction finders, and barracks for 
the personnel to man them. In September 
1 94 1, 1,500,000 yen ($351,600) was 
devoted to building military barracks, 
baths and latrines, kitchens, infirmaries, 
storehouses, workshops, torpedo storage 
sheds, garages, and air raid shelters. The 
order authorizing this expenditure specif- 
ically stated that these structures were 
intended for the use of a base force and a 
defense force, both of which under Japa- 
nese naval organization were acknowledged 
combat units. 5 

B For similar details on Japanese prewar mili- 
tary preparations in the Marshalls, see Growl and 
Love, Gilberts and Marshalls, Chapter IV. 



There is other evidence to prove that in 
the year or two before Pearl Harbor Japan 
was making active preparations to use the 
mandates, including the Marianas, as mil- 
itary and naval bases, contrary to the terms 
of the League Covenant. 

On 15 November 1939 the 4th Fleet of 
the Imperial Japanese Navy was organized 
and placed in charge of garrisons and for- 
tifications in the mandates. ' This fleet, 
which was primarily a base defense unit 
rather than the more orthodox type of 
naval combat unit, established its head- 
quarters at Truk. For administrative and 
defense purposes, the mandates were di- 
vided into four sectors — East Carolines, 
West Carolines, Marshalls, and Marianas, 
with sector headquarters at Truk, the 
Palaus, Kwajalein, and Saipan, respective- 
ly. Each sector was controlled by a base 
force subordinate to the 4th Fleet, and each 
base force commanded subordinate shore 
and surface units within its own sector. 

Shortly after the creation of the 4th 
Fleet, the 5th Special Base Force was acti- 
vated in Japan and assigned the duty of 
preparing for the fortification and defense 
of the Marianas. Attached to it were the 
5th Communications Unit and the $th De- 
fense Force, the latter unit comprising the 
bulk of the combat personnel located in 
this area before the outbreak of war with 
the United States. These troops arrived in 
the Marianas in December of 1940, Not 
long afterward their strength was augmen- 
ted by a detachment each of the 4th Fleet 
Naval Stores Department and the 4th Nav- 
al Air Depot, both located on Saipan. 7 

fi USSBS (Pacific), Naval Analysis Division, 
The Reduction of Truk (Washington, 1947), p. 2. 

7 Base Forces and Defense Forces, Early Scries, 
Vols. 9 and 10, NA 12245 ant ^ NA 13229, WDC 
106869 and WDC 160867; Special Forces, Early 

The mission of the 5th Special Base 
Force from December 1940 to 31 May 
1 94 1 was to defend its assigned areas and 
speed up preparations for combat in the 
event of a war. The subordinate §th De- 
fense Force engaged in construction of gun 
positions, road building, harbor improve- 
ment, and sundry other duties aimed at 
enhancing the security of Saipan. During 
the succeeding period, from 1 June through 
30 November, an additional mission was 
assigned to the 5th Special Base Force- - 
that of "planning and preparation for the 
Guam invasion operation." No additional 
evidence need be adduced to show that 
well before the Pearl Harbor attack, Japan 
had committed herself unequivocally to a 
policy of fortifying the Marianas for offen- 
sive as well as defensive purposes. 

From Pearl Harbor to Invasion 

For the first two years after the Pearl 
Harbor attack, the fighting war between 
Japan and the United States remained far 
from the shores of Saipan and her sister 
islands except, of course, for the Japanese 
invasion of Guam. The Marianas during 
this period served the Japanese chiefly as 
supply and staging bases for troops, ships, 
and planes engaged in battle well to the 
east and south, and the strength of combat 
naval shore units in the area remained low. 
The 5th Special Base Force on Saipan 
ranged from a low of 919 military troops 
and 220 civilians in May 1943 to a total of 
1,437 men in February of the following 
year. The 54th Naval Guard Force on 
Guam had 302 men in September 1942, 
and in early 1944 received an additional 
425 recruits. In September 1943 the Yoko- 

Series, Vols. 9 and io, NA 12226 and NA 12255, 
WDC 160867 and WDC 161009. 



suka ist Special Naval Landing Force, 
numbering about 1,500 men, arrived on 
Saipan, but in the following January was 
reduced by about a third when detachments 
were sent to the Rabaul area. 8 

Considering the magnitude of the Em- 
pire's troop commitments elsewhere, the 
garrison assigned to the Marianas can be 
considered no more than a token force. 9 
This is not at all surprising. The Marianas 
were a rear area. Simple military logic dic- 
tated that the Japanese concentrate their 
efforts in the Rabaul-New Guinea area to 
the south and build up the defenses of the 
Gilberts and Marshalls to the east. There 
was not enough money, manpower, or ma- 
teriel to build strong fortifications and de- 
fenses on every one of Japan's myriad 
island possessions, and those farther away 
from the direct line of American advance 
had to suffer neglect. Not until the Mar- 
shall Islands finally fell to the Americans 
did the position of the Marianas become 
dangerous enough to justify urgent meas- 
ures in their defense. 

Reinforcement of the Marianas 

By February 1944 the Marianas garri- 
sons could predict that their time had 

s JICPOA Trans 4071, Monthly Personnel To- 
tals for Units Under the Jurisdiction of the 4th 
Fleet, September 1942-July 1943; CINCPAC- 
CINCPOA Trans 11601, 5th Special Base Force 
(Saipan), Situation Rpt, dated 12 Feb 44; 
CINCPAC-CINCPOA Trans iao6o, account of 
the civil and military situation on Guam from the 
Japanese occupation until February 1944; Tabu- 
lar Records of Special Landing Forces, NA 11651, 
WDC 161406. 

For example, in January of 1944 the number 
of Japanese Army personnel in the Marshalls, on 
Wake, and on Kusaic numbered 13,721. This does 
not include the numerous naval personnel stationed 
in the same area whose exact numerical strength 
is not known. See Crowl and Love, Gilberts and 
Marshalls, p. 210. 

come. Tarawa and Makin in the Gilberts 
had been captured by U.S. forces in 
November 1943. Kwajalein Atoll in the 
Marshalls fell in early February 1944, and 
Eniwetok, less than a thousand nautical 
miles from Saipan, in mid-February. 10 
Also in mid-February, Admiral Mitscher's 
Fast Carrier Task Force (Task Force 58) 
executed a two-day raid against Truk, thus 
opening the way to the complete neutrali- 
zation of that formidable bastion. 11 Fol- 
lowing the Truk raid, Mitscher moved on 
to the Marianas and on 22-23 February 
administered to those islands their baptism 
of fire. 12 The rear area had obviously be- 
come a forward area. 

As the first step in recognition of the 
approaching threat to the Marianas, Caro- 
lines, and Palaus, the Japanese high com- 
mand reorganized the command structure 
in the Central Pacific. For the first two 
years of the war, the 4th Fleet had com- 
manded all Japanese garrisons in the 
mandates and was itself directly responsible 
to the Combined Fleet. By March of 1944 
the 4th Fleet had lost effective control of 
its remaining garrisons in the Marshalls 
and had been further weakened by Mits- 
cher's February raids on Truk and the 
Marianas. On 1 o March 1 944 a new head- 
quarters was placed between the Combined 
Fleet and the 4th Fleet and given control 
of all Navy and Army forces in the 
mandates and in the Bonin Islands to the 
north. This was the Central Pacific Area 
Fleet, commanded by Vice Adm. Chuichi 
Nagumo, who had led the attack on Pearl 
Harbor. Thereafter, 4th Fleet control was 
confined to naval garrisons in Truk and the 

10 Crowl and Love, Gilberts and Marshalls. 
11 Morison, Aleutians, Gilberts and Marshalls, 

12 See below, p. 71. 



eastern Carolines, though it was theoreti- 
cally exercised over the lost garrisons in 
the Marshalls. Naval garrisons in the 
Marianas, Bonins, and western Carolines 
(Palau Sector) fell under the direct con- 
trol of the Central Pacific Area Fleet. Also, 
this new headquarters, in theory at least, 
commanded all Army forces in the man- 
dates and the Bonins, commanding through 
Headquarters, 31st Army, In fact, Army 
troops remained practically independent of 
the naval command, and to all intents and 
purposes 31st Army had exclusive tactical 
and administrative control over all Army 
personnel in the area. Speaking of the 
Central Pacific Area Fleet headquarters, a 
Japanese naval commander captured on 
Saipan said, "It is a purely administrative 
command and has no tactical signifi- 
cance. ... in actual fact it never got 
beyond the stage of paper organization." 18 
Failure to establish clear-cut command 
relationships between the Army and Navy 
was characteristic of Japanese military or- 
ganization in the Central Pacific, In the 
Marianas, as elsewhere, it had serious con- 
sequences. From the very beginning there 
was friction between Army and Navy from 
the highest headquarters to the lowest 
ranks. In early March, about the time the 
Central Pacific Area Fleet was officially 
placed in over-all command, a furious ad- 
ministrative squabble arose between that 
headquarters and the 31st Army, the latter 
objecting to being subordinate to the Navy. 
The final decision, reached on 15 March, 
took the form of a compromise between 
the two headquarters. It was agreed that 
the command of each separate island was 

to rest with the senior Army or Navy of- 
ficer present. It was also agreed orally be- 
tween the Commander in Chief, Central 
Pacific Area Fleet, and the Commanding 
General, 31st Army, that neither would 
assume complete responsibility, thus ap- 
parently leaving the area without a su- 
preme command. 14 

The failure to carry out the principle of 
a unified command was to prove seriously 
detrimental to the efficiency of future 
Japanese operations in the Central Pacific. 
The resultant confusion was further com- 
pounded because of the high degree of 
mutual interdependence that necessarily 
existed between Army and Navy units. For 
example, all of the air strength in the Cen- 
tral Pacific was naval, under command of 
the Central Pacific Area Fleet, However, 
many of the air installations were serviced 
by Army units. Similarly, although the 
Army was only partially dependent on the 
Navy for surface transport, Army convoys 
had to be escorted by Navy ships. 15 

More significant than the administrative 
changes reorganizing the structure of com- 
mand in the Central Pacific was the rapid 
acceleration of troop movements into the 
area following the fall of the Marshalls and 
the strike against Truk. By May 1944 the 
Japanese had five divisions, six independent 
brigades, and five independent regiments 
in the 31st Army area, supported by in- 
numerable smaller units ranging in size 

l;) NTLF Rpt Marianas, Phase I, G-a Rpt, p. 
98, The organization of the Central Pacific Area 
Fleet is shown in Central Pacific Area Fleet Or- 
ders, NA 11810, WDC i;,094i. 

11 CINCPAC-CINCPOA Trans 1H058, Excerpts 
Taken From a StalT Diary of the 31st Army 
Headquarters, 25 February-3 1 March 1944, in 
CINCPAC-CINCPOA Trans, Vol. :i , pp. 99-148. 

15 Japanese Studies in World War II, 72, His- 
tory of the Army Section, Imperial General 
Headquarters, 1 941- 1945, pp. 94-95, OCMH; 
CINCPAC-CINCPOA Trans 10145, 31st Army 
Staff, table showing passengers and cargo of the 
Natxu 2 Convoy, dated 18 Mar 44, in CINCPAC- 
CINCPOA Trans, Vol. 3, pp. 79-98. 



from tank and antiaircraft artillery regi- 
ments down to independent machine can- 
non companies. About one third of the 
Army personnel in the Central Pacific was 
concentrated in the Marianas, including 
two divisions, two independent brigades, 
and two independent regiments. 18 The two 
divisions sent to the Marianas were the 
2Qth and the 43d. The agth was trans- 
ferred from Manchuria to Saipan in Febru- 
ary, later moving to Guam. The 43d, 
organized in June 1943, moved from Japan 
to Saipan in late May 1944. 17 

In addition, the Japanese organized sun- 
dry independent Army units for service in 
the Central Pacific Area. Units of battalion 
size and smaller were detached from their 
parent divisions and reorganized into eight 
expeditionary units, three of which were 
sent to the Marianas. The 1st Expedition- 
ary Unit, consisting of four infantry bat- 
talions and two artillery battalions, was 
allocated to Saipan; the 5th Expeditionary 
Unit, of two infantry battalions and one 
artillery battalion, was moved to Pagan; 
and the 6th Expeditionary Unit, made up 
of six infantry battalions and two artillery 
battalions, was sent to Guam. 18 

In May, after most of the expeditionary 
units had reached their destinations, Army 
Section, Imperial General Headquarters, 
ordered a reorganization of the expedition- 
ary units into independent mixed brigades 
and independent mixed regiments. In the 
Marianas the 1st Expeditionary Unit (Sai- 
pan) became the 47th Independent Mixed 
Brigade, the 5th Expeditionary Unit (Pa- 

16 Japanese Studies in World War II, 55, Oper- 
ations in the Central Pacific, pp. 17-19, 36-38, 

17 MID WD, Order of Battle of the Japanese 
Armed Forces, i March 1945, pp. 99-1OO; Japa- 
nese Studies in World War II, 72, pp. 92-93. 

18 Japanese Studies in World War II, 55, p. 17. 

gan) became the gth Independent Mixed 
Regiment, and the 6th Expeditionary Unit 
(Guam) was divided into the 48th Inde- 
pendent Mixed Brigade and the 10th In- 
dependent Mixed Regiment.™ 

The Navy, too, began to increase its 
strength in the Marianas shortly after the 
fall of the Marshalls, The 55th and 65th 
Naval Guard Forces were established on 
Saipan and Tinian, respectively, and sever- 
al antiaircraft artillery units were also 
dispatched to the area. A large variety of 
small administrative offices were established 
to handle the greatly increased volume of 
supplies and troop movements, and labor 
units for construction work and loading 
and unloading ships were added to the gar- 
risons. Finally, air strength was poured into 
the Marianas during the period as local 
airfields were developed and war came 
closer, so that many naval airmen and 
maintenance men swelled the number of 
naval air personnel in the islands. 

This accelerated movement of troops into 
the Marianas was not allowed to go un- 
contested by the ever-advancing American 
forces. From the very beginning of the Jap- 
anese attempt to reinforce the islands, 
submarines of the U.S. Pacific Fleet began 
to take their toll. American submariners 
played an important role in the capture of 
the Marianas; in fact it can be said that 
the American campaign actually began in 
February with the first submarine attacks 
on Japanese troop convoys bound for 

The first of the major Japanese troop 
movements involved the transfer of the 
agth Division, which left Ujina, Japan, 

16 CINCPAC-GINCPOA Trans 16035, cha « in- 
dicating change in organization of old units 
(undated), rev. trans by Hq NTLF in the field, 23 
Jul 4-4- 



on 26 February aboard three troop trans- 
ports. Late in the afternoon of the 29th 
the convoy was attacked by American sub- 
marines. One transport {Sakito Maru), 
laden with 3,080 troops was torpedoed and 
sunk. Only 1,688 were rescued. The sur- 
vivors (members of the 18th Infantry Reg- 
iment ) landed on Saipan with almost no 
equipment. According to one report, "All 
their weapons were lost except seven rifles, 
one grenade thrower, two light machine 
guns and 150 bayonets." 20 

The next large convoy to sail for the 
Marianas left Yokohama on 1 2 March and 
carried the 1st, $th, and 6th Expeditionary 
Units. The convoy was attacked by Ameri- 
can submarines and, although no Army 
troops were lost, a naval transport (Ko~ 
kuyo Maru) carrying 1,029 reinforcements 
for the 54th Naval Guard Force on Guam 
was torpedoed and sunk. 21 

In April two more convoys left Japan 
for the Marianas, and although the first 
was attacked by submarines and two of its 
ships sunk, all troops were rescued and put 
ashore safely on Saipan. 22 The last two 
convoys to reach the Marianas, however, 
arrived with the units seriously depleted 
and without their equipment. The first of 
these departed Yokohama on 5 May and 
reached Saipan on the 14th. None of the 
troops carried aboard were originally in- 

3(1 CINCPAC-CINCPOA Trans 12058. 

2 L CINCPAC-CINCPOA, Trans 10145, PP- 79- 
98; CINCPAC-CINCPOA Trans 12058, pp. 34- 
36; CINCPAC-CINCPOA Trans 9767, Situation 
Rpt, 4th Special Shipping Engineer Company 
(Saipan), dated Apr 44. 

22 CINCPAC-CINCPOA Trans 10931, a file of 
orders and tables showing troop movements and 
locations of units in the Central Pacific Area. The 
Joint Army-Navy Assessment Committee, Japanese 
Naval and Merchant Shipping Losses During 
World War II By All Causes (Washington, 1947) 
(hereafter cited as JANAC, Japanese Shipping 
Losses) . 

tended for the Marianas, but when two of 
the transports were torpedoed the survivors 
were landed on Saipan. Fifteen hundred 
of these, members of the gth Expeditionary 
Unit bound for Yap, remained on Saipan 
until the American invasion. About six 
hundred were reorganized as a battalion of 
the 47th Independent Mixed Brigade, but 
the rest remained on Saipan as ill-equipped 
stragglers. Other survivors of the same ill- 
fated convoy belonged to the 15th Infantry 
Regiment, destined for the Palaus. They 
formed another straggler group on Saipan, 
where the American invasion caught 
them still awaiting transportation to the 
Palaus. 23 

The last major troop movement to the 
Marianas, and certainly one of the most 
significant, was the transfer of the 43d Di- 
vision from Japan to Saipan. Arriving only 
a few weeks before the American invasion, 
the division was to play a leading role in 
the defense of the island, and its com- 
mander, Lt. Gen. Yoshitsugu Saito, would 
assume effective control of the Saipan 
defenses. For transportation purposes the 
division was divided into two echelons, the 
first of which made its way safely through 
to Saipan sometime in May. The second 
was not so fortunate. On 30 May a convoy 
of seven transports carrying more than 
7,000 troops of the 43d Division sailed 
from Tateyama and headed south. It was 
subjected to almost continous submarine 
attack, and within three days five of the 
seven transports were sunk. The two re- 
maining vessels picked up the survivors and 

23 CINCPAC-CINCPOA Trans 1093 1 ; 
CINCPAC-CINCPOA Trans 9485, Army Section 
Order 982 (1 Apr 44); CINCPAC-CINCPOA 
Trans 10959, Extracts From an Account of the 
Sinking of the Hiyori Maru; CINCPAC- 
CINCPOA Trans 9883, File of Shipping Orders, 
Shipping Operation Order 283 (7 Apr 44). 



steamed the rest of the way to Saipan. 24 
About 80 percent of the troops of this 
convoy were saved and landed on Saipan, 
but they arrived as hapless survivors with- 
out weapons or equipment. The 11 8th In- 
fantry Regiment lost about 850 men, and 
the survivors had virtually no resemblance 
to the organized fighting team that had 
left Japan. So little time was left before 
the American invasion that the regiment 
could not be reorganized and re-equipped 
sufficiently to raise its combat efficiency 
much above nil. 25 

All together, from January to early 
June 1944, the Japanese dispatched about 
45,000 Army troops to the Marianas. Of 
these about 40,000 were allotted to Saipan, 
Guam, and Tinian; the remainder to Pagan 
and Rota. At least 12,000 of these troops 
were aboard torpedoed vessels, and about 
3,600 died as a result of the sinkings. While 
many of the survivors were successfully re- 
organized, rehabilitated, and re-equipped, 
about half, perhaps four or five thousand, 
became stragglers on Saipan, equipped and 
armed only with their resolution to die for 
the Emperor, Thus, well before the initial 
American strikes against Saipan from sur- 
face ships and aircraft, U.S. submarines 
had seriously disrupted Japan's major ef- 
fort to reinforce the Marianas against the 
imminent threat of invasion. 

Military Construction in IQ44 

Along with their hasty and not al- 
together successful effort to enlarge the 
garrisons of the Marianas, the Japanese 
in early 1944 undertook an ambitious 
program of building up the islands' fortifi- 

21 USSBS, Interrogations I, 31a; CINCPAC- 
GINGPOA Trans 1093 1, p. 40. 

' i! > C1NCPAC-CINCPOA Tram 10931, p. 40. 

cations and defenses. First priority was as- 
signed to airfield construction. At the be- 
ginning of the war Saipan had an airfield 
(Aslito) and a seaplane base and Tinian 
had an airfield. All three appear to have 
been operational. The situation remained 
unchanged as late as mid- 1943, when the 
Imperial Navy commenced a new program 
of airfield construction in the Marianas 
and Carolines, planning to increase the op- 
erational airfields from one to two on Sai- 
pan and from one to three on Tinian. In 
addition, the program called for two new 
fields on Guam, and one each on Pagan 
and Rota, where none had existed before. 
By February 1944 Aslito airfield was being 
enlarged, and two more fields on Saipan 
were under construction. The Saipan sea- 
plane base, constructed in 1935, was in full 
operation. On Tinian, the Ushi Point field 
was in operation, and another field was 
under construction. On Guam, Sumay air- 
field was almost completed, and three other 
fields and a seaplane base were under way 
or in the planning stage. 2 * 5 

In March plans were developed for an 
even more rapid build-up of airfield facil- 
ities. Combined Fleet and Central Pacific 
Area Fleet each issued orders outlining an 
ambitious policy of airfield construction. 
The Combined Fleet order provided that 
three independent complexes of bases were 
to be rapidly completed in the Marianas- 
Tnik area and in the Palau-Yap area. 27 
The Central Pacific Area Fleet order de- 

' 2,i JICPOA Trtms 5577, nth Air Fleet Secret 
Bull 37, data tables for surface Craft assigned or 
attached to the illh Air Fleet; Japanese Studies 
in World War II, Go, The AGO Operation, 1944, 
10092, Advance Expeditionary Unit, 13th Divi- 
sion, intelligence record dated 29 February 1944. 

27 CINCPAC-CINCPOA Trans 9497, Combined 
Fleet Secret Order 20 (19 Mar 4,4), Construction 
of Air Bases in the Central Pacific Area, 



fined construction policy as: "To build 
rapidly a large number of bases so as to 
make possible the immediate development 
of great aerial strength." In the Marianas, 
the bases were first to be built on a 
rough and ready basis and then gradually 
brought to a finished state as conditions 
permitted. Each field was to be integrated 
into the land defenses; machine gun po- 
sitions for plane-mounted guns were to be 
built near plane shelters, and air force per- 
sonnel were to be organized for ground 
combat. 28 

These plans were intended to result in 
fourteen airfields and two seaplane bases: 
Saipan, three airfields and a seaplane base; 
Tinian, four airfields; Guam, four airfields 
and a seaplane base; Pagan, two airfields; 
Rota, one airfield. Each airfield was to be 
capable of handling forty-eight planes, ex- 
cept for the Marpi Point field on Saipan, 
which would handle twenty-four, and the 
Charan Kanoa strip, also on Saipan, which 
was for emergency use only. The entire net- 
work of bases would be sufficient for six 
hundred planes of various types. 2 " 

Although the Combined Fleet order spec- 
ified that normal air facilities were to be 
completed by April, the estimate proved 
much too optimistic. When the American 
landing forces came upon the scene in 
June, much of the construction was still 
unfinished. On Saipan, Aslito airfield was, 
of course, operational, as was the Charan 
Kanoa emergency airstrip, but the Marpi 
Point field was still unfinished and non- 
operational. Guam had two operational 

2S CINCPAC-CINCPOA Trans 949H, Central 
Pacific Area Force Secret Order 2 (21 Mar 44), 
Construe lion of Air Bases in the Marianas and 

2 * Ibid.; CINCPAC-CINCPOA Trans 9692, 
Central Pacific Area Fleet Secret Order 30-8 (ao 
Apr 44), New Nomenclature for Air Bases. 

strips and two still unfinished. Tinian, the 
air center of the Marianas, had three fields 
finished and one to go. Rota and Pagan 
each boasted a surfaced runway. All to- 
gether, of the fourteen fields planned, 
nine had been completed and were capable 
of handling at the most a total of four 
hundred planes rather than the originally 
planned six hundred. 30 

While the Imperial Navy was hurriedly 
attempting to build up air strength in the 
Marianas, 31st Army was making similar 
plans for ground fortifications. The precis 
of the 31st Army defense plan gives a gen- 
eral outline of the Army fortification pro- 
gram : 

While deploying the troops the defensive 
constructions must be strengthened and gen- 
eral preparations for the annihilation of the 
enemy landing on the beaches must be com- 
pleted, including the protection of our air 
bases. The field positions must be completed 
within one month after the arrival of troops 
and within three months the positions must 
be strengthened by constructing permanent 
fortifications in the most important points. 31 

The hope that this schedule could reach 
completion before the American landings, 
if ever seriously held, proved a vain one. 
The 4.3d Division did not arrive on Saipan 
until early June, a matter of days before 
the invasion. The jth and 16th Indepen- 
dent Engineer Regiments, which were re- 
sponsible for most of the "permanent 
fortifications," arrived in April, only two 
months before the invasion. The job, even 
under ideal conditions, would have been 
difficult, and conditions were far from 
ideal. Troops arrived in a haphazard fash- 
ion, often depicted in number and missing 

30 TF 56 Rpt Fokagkr, G-2 Rpt, App. F, pp. 

31 Japanese Studies in World War II, 55, pp. 



their arms and equipment. Moreover, 
American submarine warfare accounted for 
the loss of more than mere manpower. Es- 
sential building materiel went to the bot- 
tom of the sea along with the men. A 
report, dated 31 May, from the chief of 
staff of the gist Army gives a clear survey 
of the difficulties facing the Japanese in 
their hurried effort to construct adequate 
fortifications on Saipan and the other is- 
lands in the Marianas: 

We cannot strengthen the fortifications ap- 
preciably now unless we can get materials 
suitable for permanent construction. Spe- 
cifically, unless the units are supplied with 
cement, steel reinforcements for cement, 
barbed wire, lumber, etc., which cannot be 
obtained in these islands, no matter how 
many soldiers there are they can do nothing 
in regard to fortifications but sit around with 
their arms folded, and the situation is 

I would like this matter of supply of con- 
struction materials dealt with immediately. 32 

Japanese Doctrine for Island Defense 

The failure on the part of the Japanese 
to meet their time schedule in reinforcing 
the Marianas' physical defenses had an im- 
portant effect on the tactical doctrine to 
which they adhered throughout the cam- 
paign. It was one of the factors that 
compelled them to rely more heavily on 
beach defenses than would logically have 
been called for by the size and physio- 
graphic features of these particular islands. 

In the Gilberts and Marshalls, Japanese 
defensive doctrine stressed defense at the 
beaches. At Tarawa, Makin, Kwajalein, 

J3 Report on the Defenses of the Various Is- 
lands, from CofS 31st Army to CofS Central Pa- 
cific Area Fleet, Trans I iG of Japanese Document 
by Hq NTLF, 1 1 Jul 44, copy in archives of Hist 
Br G-3, Hq U.S. Marine Corps. 

and Eniwetok the fortifications were con- 
centrated on a thin line along the shores 
with little defense in depth." 3 By concen- 
trating the majority of their positions on 
the beaches proper, and with the aid of 
fringing coral reefs and offshore obstacles, 
the Japanese hoped to annihilate the Amer- 
icans before they could gain a foothold 
ashore. If the U.S. troops did succeed in 
establishing a beachhead, doctrine called 
for a counterattack to push them back into 
the sea. 

This doctrine of island defense was based 
to a great extent on purely geographical 
and terrain considerations. The Gilberts 
and Marshalls are composed of widely scat- 
tered coral atolls, each in turn composed 
of many small islets. Although the Japanese 
usually selected the larger of the islets for 
their bases, these still had very little area 
and were generally elongated in shape. De- 
fense in depth was impractical not only 
because there was little depth to defend 
but also because what little there was us- 
ually contained an airstrip that had to be 
left free of obstacles if it was to land and 
dispatch planes. Moreover, the flat terrain 
of the atolls provided no natural features 
such as hills and caves that could be ex- 
ploited to set up an adequate defense in 

The Marianas are much different. They 
are volcanic islands, not coral atolls. They 
arc generally much larger in size, have con- 
siderable elevation, and the terrain is rug- 
ged and mountainous, providing favorable 
opportunities for defense in depth. Yet, in 
spite of this the Japanese continued to 
place great emphasis on defending the 
shore lines of Saipan, Tinian, and Guam 

33 See Growl and Love, Gilberts and Marshalls, 
Chs. IV, XIII. 



to the consequent neglect of fortifications 
and gun positions in the interior. 

The continued reliance on beach de- 
fenses is illustrated in the defense plan for 
Saipan drawn up by the ist Expeditionary 

Tactical Command Doctrine 
A, Objective 

It is expected that the enemy will be de- 
stroyed on the beaches through a policy of 
tactical command based on aggressiveness, 
determination, and initiative. . . , 

When the enemy elements are attempting 
to land : The main fire-power will be directed 
at the enemy landing forces prior to their 
arrival on the beach. Taking advantage of 
the confusion, the enemy will be rapidly de- 
stroyed by counter attacks, mounted from all 
sectors wherever the opportunity presents 

Should the enemy succeed in gaining a 
foothold on the beach, intense fire will be 
concentrated and determined counter-attacks 
launched with the aid of reserves and tanks. 
Although the advantages of surprise will be 
lost, the enemy landing forces can be dealt 
with by further attacks after night fall. 84 

Later in the war and under conditions 
similar to those obtaining in the. Marianas, 
the Japanese abandoned or modified their 
earlier doctrine and concentrated heavily 
on defenses in depth. When the U.S. Ma- 
rines landed on I wo Jim a, they found a 
well-prepared network of defenses in depth 
as well as fortifications commanding the 
shore line.'"' 3 At Okinawa the landing 
beaches on the west coast were left prac- 

:M CINCPAC-CINCPOA Trims 8946, 1st Expe- 
ditionary Unit Defense Plan for Saipan Island, 
dated 10 May 44. 

3r ' Isely and Crowl, The U.S. Marines and Am- 
phibious War, Ch. X.; Lt. Col. Whitman S. 
Bartley, USMC, Iwo Jima: Amphibious Epic, 
Historical Branch, G-3 Division, Headquarters 
U.S. Marine Corps (Washington, 1954), pp. 5-18. 

tically undefended, and the main Japanese 
forces had retired before the invasion to the 
southern part of the island, where they 
holed in along the Shuri line with its elab- 
orate system of caves and underground 
installations. 38 

Why, then, did the Japanese as late as 
June 1944 cling to the older concept, which 
though valid enough for atoll warfare was 
clearly not so suitable for the type of fight- 
ing that would inevitably develop in the 
mountainous terrain of the Marianas? The 
failure to adjust tactical doctrine to chang- 
ing conditions of terrain can probably be 
attributed in part to the highly aggressive 
spirit of the Japanese military mind. Gen- 
erally, the Japanese preferred to sally forth 
sword in hand against the enemy rather 
than bide their time in prepared positions. 
As noted in the U.S. Army handbook on 
the subject of Japanese tactics: 

No matter what the situation, a Japanese 
commander's first reaction to it is to act ag- 
gressively to maintain the traditions of his 
army. . . . Even when the Japanese com- 
mander assumes the defensive, he will, so far 
as possible, carry out that defense by using 
the most aggressive tactics that the situation 
permits. 37 

In the case of the Marianas, there was an 
even more compelling consideration that 
forced the Japanese to rely most heavily on 
their beach defenses. There was simply not 
enough time to complete the fortification 
program. The 31st Army program for de- 
fense had made provision for falling back 
upon prepared "strategic inland positions" 

3 " Roy E. Applcman, James M. Burns, Russell 
A. Gugclcr, and John Stevens, Okinawa: The Last 
WAR II (Washington, 1948). 

:l7 TM-E 30-480, Handbook on Japanese Mil- 
itary Forces, Ch. VII, Japanese Tactics, 1 Jun 
4 .1. P- 2 - 



in the event that an enemy landing force 
was not thrust back into the sea. That plan 
reads, in part: 

First priority in construction will be given 
to improvised positions designed to frustrate 
enemy landings on the beaches . . . and to 
temporary protective measures designed to 
minimize our losses in personnel and materi- 
als. Later, these constructions will be rapidly 
supplemented and strengthened by extending 
the positions in depth, converting actually 
the island into an invulnerable fortress. ,ss 

The speed-up of the U.S. invasion plans, 
coupled with the loss of valuable building 
materials to the U.S. submarines, made the 
fulfillment of the second part of this plan 

In line with the doctrine of defense at 
the beaches, the 31st Army planned to 
cover all segments of the shore line "where 
the enemy could land without difficulty" 
with independent strongpoints several hun- 
dred meters or one kilometer apart. Each 
strongpoint would be manned by an in- 
fantry company or a heavy weapons pla- 
toon. Patterns of fire were to be arranged 
so that each strongpoint could provide 
flanking fire on its neighbors' fronts, and 
areas not otherwise provided for were to 
be covered by machine guns and mor- 
tars. 39 In this manner, whole companies 
would be deployed along the beaches. 

Behind the coastal positions, at a dis- 
tance depending on terrain, was to be 
constructed a second line that would cover 
any partial collapse of the coastal positions 
and serve as a starting point for counter- 
attacks by reserve units. The second line 
was to be organized on the same principle 
as the coastal positions. The 31st Army 
plan emphasized the construction of dum- 

my positions between strongpoints and be- 
tween the coastal positions and the second 
line. These positions were to deceive the 
enemy and divert his fire. They were to be 
especially thick on those stretches of the 
shore line where enemy landings were less 
likely because of natural obstacles, and 
where only small forces or lookouts might 
be stationed. 

Since the artillery was to fire antiboat 
missions, many pieces were to be emplaced 
in the coastal positions and the second 
lines, as well as farther to the rear. Anti- 
aircraft artillery was to be located in po- 
sitions where it could lend support to 
ground actions as well as fulfill its primary 
function. Finally, the plan made rather 
vague provision for final strongpoints to 
the rear of the second lines, providing that 
"if time allows the rear positions . . . 
must be strongly built and also completely 
equipped for the counter-attack." 40 Ac- 
tually, time did not allow the Japanese to 
provide for such positions. 

Enemy Troop Strength and 
Dispositions on Saipan 

Command of all Army troops in the 
Marianas rested with the 31st Army under 
Lt. Gen. Hideyoshi Obata, who had head- 
quarters on Saipan. 41 When the American 
forces landed on 15 June, Obata was in 
the Palau Islands and consequently exer- 
cised no tactical command during the 
campaign. His headquarters on Saipan, 
consisting of about 1,100 officers and men, 
was largely administrative in function and 
had little tactical significance. The largest 
single Army unit on the island at the time 

?fl Studies in World War II, 55, p. 25. 
39 Ibid., pp. 26-32. 

' 10 Ibid., p. 29. 

:l] See below, App. C, Japanese Order of Battle 

on Saipan. 



of the landing was the 43d Division, com- 
manded by General Saito. It was Saito who 
actually exercised tactical command until 
his death a few days before the close of the 
battle. The division consisted of three in- 
fantry regiments, the 118th, 135th, and 
136th, plus a signal company, a transport 
company, an ordnance company, a field 
hospital, and an "intendance duty unit" 
responsible for quartermaster and finance 
functions. All together, the division num- 
bered about 12,939 officers and men. 43 

Next in size among Army units was the 
47th Independent Mixed Brigade, com- 
manded by Col. Yoshiro Oka. The brigade 
had four organic battalions, but one, the 
315th, was on Pagan. The three others, 
the 316th, 317th, and 318th Independent 
Infantry Battalions, were on Saipan. An 
engineer unit and an artillery unit, the lat- 
ter consisting of one battery of eight field 
guns and two batteries of seven howitzers 
each, also belonged to the brigade. Total 
strength of the brigade on Saipan at the 
time of the landing was about 2,600. 

To this nucleus was attached a host of 
smaller units including many that had orig- 
inally been destined for other islands but 
had been stranded on Saipan as a result 
of shipping damage inflicted by American 
submarines. The most important of these 
were the 3d Independent Mountain Artil- 
lery Regiment consisting of two battalions, 
each with twelve 75-rnni. mountain guns; 
the 1 6th Shipping Engineers Regiment; the 
7th Independent Engineer Regiment; the 
gth Tank Regiment with thirty-six me- 
dium and twelve light tanks; and the 
35th Antiaircraft Artillery Regiment. Total 
strength of Japanese Army troops on Sai- 

12 Sources for this data on the Japanese Order 
if Buttle are given in Appendix C. 

pan on the eve of the American invasion 
was about 25,469. 

To this number must be added about 
6,160 naval personnel. Among the naval 
units present on the island were the Head- 
quarters, Central Pacific Area Fleet, under 
Admiral Nagumo and the Headquarters, 
5th Base Force, under Rear Adm. Takahisa 
Tsujimura. The latter unit had been in 
command of the Marianas since before 
Pearl Harbor, but with the arrival of Army 
troops in 1944 it assumed a less important 
role in the defense of Saipan, though it 
continued to command naval shore forces 
and surface units within the Marianas. The 
largest single element of the naval forces 
was the 55th Naval Guard Force, about 
2,000 officers and men, which was chiefly 
responsible for manning coast defense guns. 
The only other large naval unit was the 
Yokosuka 1st Special Naval Landing Force, 
consisting mainly of a headquarters, three 
rifle companies, a machine gun platoon, 
and a gun section. 

For purposes of ground defense. Sa ipan 
was divided into four parts.l (A/fl/? 3) The 
first was the Northern Sector, tnat portion 
of the island north of a line beginning at 
Flores Point on the west coast and extend- 
ing to the cast coast in a southeasterly di- 
rection. The most important troop unit 
located here was the 135th Infantry Regi- 
ment, less one battalion. 

The second division was the Navy Sec- 
tor, which was bounded on the north by 
the Northern Sector, on the west by the 
shore line from Flores Point to a point just 
below Garapan, thence on the south by a 
line to a point just southwest of Mount 
Tapotchau's summit, and on the east by 
a line from this point up the axis of the 
island to the center of the Northern Sector's 
boundary. The §th Base Force controlled 

Marpi Pt 

F lores Pt 


Afetna P* 
Char an Kanoa 




^^^^— Sector boundary 
[~] japanese units 
SNLF Special Naval Lanoing Force 

MAP 3 

F. Tempt* 



the Navy Sector, but the beach defenses 
were actually manned by a battalion of the 
1 36th Infantry Regiment. 

The Central Sector lay directly south of 
the Navy Sector. Its shore line extended 
from a point below Garapan to Afetna 
Point. Inland, to the east, it was bounded 
by an extension of the Navy Sector's boun- 
dary down the central axis of the island to 
a point directly east of Afetna Point. Here 
the boundary turned west to Afetna, en- 
closing the sector from the south. Responsi- 
bility for the defense of the Central Sector 
rested with the commanding officer of the 
136th Infantry Regiment. His regiment 
was reduced to the 'jd Battalion rein- 
forced by one company of the 3d, since the 
1st Battalion was assigned to the Navy 
Sector and the remaining two companies of 
the 3d Battalion were in 43d Division 

The fourth defense area was the South- 
ern Sector, whose shore line extended 
along the west coast from Afetna Point, 
around the entire south coast, and up the 
cast coast beyond Kagman Peninsula to 
the boundary line of the Northern Sector. 
This sector, commanded by the command- 
ing officer of the 4Jth Independent Mixed 
Brigade, encompassed about half of the is- 
land, and probably contained the same 
proportion of the island's troops. Here were 
concentrated all the reserve troops as well 
as the forty-eight tanks, the main artillery 
batteries, and the antiaircraft units clus- 
tered around Aslito airfield. 43 

13 Map, Sectors for Units Defending Saipa.ii, re- 
produced from overlay attached to captured 
Japanese Northern Marianas Group Operations 
Order A-4, reproduced in NTLF Rpt Marianas, 
Phase I, C-a Rpt, pp. 8-9; 136th Infantry Regi- 
ment Operation Order 12, 0800, 25 May, copy in 
Field Diary, ad Artillery Battery, 136th Infantry 
Regiment, 1—3 1 May 1944 (NA 21961). The di- 

Within each of these sectors, the dispo- 
sition of troops was generally in accordance 
with the doctrine of concentrating on de- 
feating the enemy at the shore line. For 
example, a 136th Infantry Regiment op- 
erations order dated 25 May makes the fol- 
lowing disposition of the regiment's forces 
in the Central Sector, which included most 
of the shore line upon which the 2d Marine 
Division was to make its landing: 44 

Front Line (Beaches) 

4th Company 

§th Company 

6th Company (less 1/3) 

8th Company (less 1/3) 

2d Infantry Gun Company 
Sector Reserves 

1/3 6th Company 

1/3 8th Company 

The commitment to beach defense of such 
a large portion of troops of the 136th Regi- 
ment with the consequent neglect of reserve 
power was entirely consistent with prevail- 
ing Japanese tactical doctrine. 

Reserves for the island were located in 
the area just north of Magicienne Bay and 
included two companies of the 3d Battal- 
ion, 136th Infantry Regiment, with two 
companies of the 3d Battalion, yth Inde- 
pendent Mixed Brigade, attached. The 
1 6th Shipping Engineers Regiment (less 
detachments) and the main body of the 

version of the 1st Battalion, 136th Regiment, to 
the Navy Sector, a last-minute change that does 
not appear in American sources, is indicated by a 
136th Infantry Regiment Order, 2000, 29 May 44, 
signed by the regimental commander, Col. Yuki- Ogawa, and found in Field Diary, 1st 
Infantry dun Company, 136th Infantry Regiment, 
7-yi May 1944, entry for 29 May (NA 22286). 
11 136th Infantry Regiment Operation Order 7, 
1:500, 2f) May 44, Field Diary, 2d Artillery Battery, 
136th Infantry R.egim,ent_, 1-3 1 May 44 (NA 
27961 ). 



gth Tank Regiment were also in this area. 
In case of landings in the Garapan- 
Tanapag sector, the tanks were to assemble 
in the mountains two miles east of Garapan 
and there prepare to counterattack. If 
landings took place in the Magicicnnc Bay 
or Charan Kanoa area, the tanks were to 
assemble about one-half mile north of Aslito 
airfield's west end and prepare for opera- 
tions against either area, or both.' 15 

Most of the Army mobile artillery, or- 
ganized as the 3d Independent Mountain 
Artillery Regiment and the 4Jth Indepen- 
dent Mixed Brigade Artillery Unit, was 
located along the ridge overlooking the 
Charan Kanoa beaches to the north and 
south of Mount Fin a Susu. From these 
positions the guns were to fire missions on 
the east and west coasts. Japanese artillery 
preparations on Saipan were careful and 
elaborate, preliminary sightings having 
been made on important points such as 
breaks in the reefs. However, these pre- 
liminary sightings seem to have been con- 
ducted independently by separate pieces of 
batteries, and no common survey control 
was established. As a result, the Japanese 
were unable to mass fires by tying in bat- 
teries or battalions, and the Japanese 
artillery dissipated itself in local un-co- 
ordinated missions. 46 Another defect in 
Japanese artillery on Saipan arose from the 
lack of prime movers. All of the field artil- 
lery was mobile, but there were neither 
horses nor trucks on the island and the 
pieces therefore had to be moved by hand, 
if at all. 

In addition to the Army's mobile artil- 
lery, Saipan could boast an elaborate 

"°NTLF Rpt Marianas, Phase I, G-a Rpt, pp. 
0, 76. 

16 CINCPAC-CINCPOA Trans 9604, Battle 
Plan of Artillery Unit of the 1 st Expeditionary 

system of naval coast defense and dual- 
purpose batteries. The Japanese Navy's 
scheme of artillery fire was intended to 
cover almost every square inch of both Sai- 
pan and neighboring Tinian, large sweeps 
of the adjacent sea, and nearly all of the 
air overhead. According to the Navy's plan, 
coast defense guns (12, 14, and 15-cm's.) 
were to be lined up along the west coast 
of Saipan from Agingan Point to Tanapag 
in a series of seven batteries covering the 
entire approach from the west with inter- 
locking sectors of fire. A second concentra- 
tion of Navy coast defense guns, in five 
batteries, was to be emplaced on Nafutan 
Point and the shores of Magicienne Bay, 
covering the sea approaches to Saipan from 
the south and southeast. A battery of 4- 
cni. guns on Mar pi Point emplaced to fire 
to the north would complete the picture. 
Two batteries of 12 -cm. dual-purpose guns 
were to cover the Aslito airfield; two more 
of the same size were to cover the center of 
island; an 8-cm. battery, the Tanapag 
area; and a battery of 12-cm's., Marpi 
Point. 47 Not all of these emplacements were 
completed at the time of the American 
landings, but most of the Navy's coast and 
aerial defense program was in working or- 
der by mid-June of 1944. 

Army antiaircraft batteries on Saipan 
were spread from Mutcho Point to Aslito 
airfield, mostly along the ridges and hills 
overlooking the west beaches. The air over 
the entire southern portion of the island 
was covered by the guns of the 25th Anti- 
aircraft Artillery Regiment's 1st, 2d, and 
Gth Batteries. The central portion of Sai- 
pan was covered by the 43d Independent 

47 CINCPAC-CINCPOA Trans 12252, Disposi- 
tion of Navy Dual-Purpose and Coastal Defense 
Batteries on Saipan and Tinian, dated 15 May 44. 
Reprinted in CINCPAC-CINCPOA Trans, Vol. 8, 
P- 9- 



Antiaircraft Company and the 45th Field 
Machine Cannon Company.^ In general, 
however, a shortage of ammunition handi- 
capped the antiaircraft artillery. Maximum 
allowance for antiaircraft weapons on Sai- 
pan was three units of fire, but in fact the 
25th Antiaircraft Artillery Regiment and 
its two attached companies probably did 
not have as much as one unit of fire (4,800 
rounds) on hand. Antiaircraft ammunition 
was in short supply not only in the Mari- 
anas but in all other 31st Army areas as 

Japanese Expectations 

Immediately after the February strike 
against Truk by Marc Mitscher's fast 
carrier task force, the Japanese reached a 
state of near panic in their apprehensions 
of an imminent American attack against 
the Marianas. Preparations for moving the 
31st Army headquarters to the Marianas 
had already begun when U.S. forces 
seized the Marshalls and Eniwctok, and the 
headquarters staff was scheduled to leave 
Tokyo for Saipan when Mitscher's follow- 
up strike on the Marianas took the Japa- 
nese by surprise. This strike occurred on 
22-23 February, and on the following day 
General Tojo himself suddenly ordered the 
staff to depart at once. On 3 March 31st 
Army issued an emergency plan for the de- 
fense of the archipelago, which at that time 
was virtually defenseless. On 4 March civil 
government was abolished on Guam, and 
shortly afterward martial law was cstab- 

18 CINCPAOCINCPOA Trans 12251, order for 
change in location of Army antiaircraft batteries 
on Saipan, dated 10 May 44. 

49 GINCPAC-GINCPOA Trans 9698, 31st Army 
(able showing immediate supply of ammunition 
necessary for antiaircraft guns and antiaircraft 
machine cannon. 

lished throughout the islands. In early 
March the Japanese actually feared an 
invasion of the Marianas by the end of the 
month, as indicated by the following ex- 
cerpt from the 31st Army staff diary: "The 
central command was very much con- 
cerned over the enemy attacks. After the 
enemy striking force hit the Marianas Sec- 
tor on the 2 2d and 23d of February, 
thereby revealing our lack of defenses, they 
were afraid of an enemy invasion in 
March." 50 The sinking of Sakito Maru, 
which was carrying the iSth Infantry Reg- 
iment to Saipan, only increased the alarm. 
For a while 31st Army headquarters 
planned to divert the 8th Expeditionary 
Unit, bound for Truk, to the Marianas, 
but this plan was not carried through. By 
late March the apprehension over an im- 
mediate invasion had somewhat abated, 
and Army troops were being used to build 
airfields. 81 

Then at the end of March Mitscher 
raided the Palaus, and on 22 April Mac- 
Arthur captured Hollandia in Netherlands 
New Guinea. The Japanese pulse once 
again beat faster. The Japanese Navy and 
Army high command differed somewhat in 
their estimates of the situation. The raid 
on the Palaus, the capture of Hollandia, 
and, finally, the operations around Biak in 
May led the naval high command to antic- 
ipate that the next American attack would 
come in the Palaus or Carolines rather than 
as far north as the Marianas. 52 On 3 May 
Admiral Soeumu Toyoda assumed com- 
mand of the Combined Fleet and received 
a directive from Imperial GHQ outlining 

50 CINCPAC-CINCPOA Trans 12038, Excerpts 
Taken From a Staff Diary of the 31st Army Head- 
quarters, 25 Fcbruary-3 1 March 1944, p. 32. 
r ' ] Ibid., pp. d, I4-I7: 32) 45- 
r>a USSBS, Campaigns, pp. 3-9, 204-15. 



the A Operation (also known as the A-GO 
Operation), which called for the surface 
striking force and land-based air units to 
be prepared for a decisive engagement with 
the U.S. fleet by the end of May, presum- 
ably in the Palau-Garolines area. 

Army officers in Imperial GHQ at the 
same time seem to have taken more serious- 
ly the possibility of an early invasion of the 
Marianas. On the assumption that the main 
objective would be the Philippines, they 
concluded that the Americans might ad- 
vance along the north coast of New Guinea 
or into the Carolines and Marianas. It was 
also estimated that "there is a great pos- 
sibility that both of the operations will be 
commenced simultaneously." 53 

Whatever may have been their belief in 
regard to an ultimate invasion of the Mari- 
anas, it is clear that the Japanese did not 
expect a full-scale attack against Saipan 
by mid- June. As late as 8 June, 31st Army 
headquarters issued an order for the trans- 
fer of troops from Saipan to other islands 
in the archipelago, which, had it been car- 
ried out, would have reduced the Saipan 
garrison by 2,500 men. 54 An even more 
striking indication that the Japanese were 
caught off guard by the rapidity of the 
American approach comes from an order 
issued by Admiral Nagumo on 14 June, 
the day before the first American troops 
hit the beaches on Saipan, that reads, in 

5il Japanese Studies in World War II, 7a, p, ia'-s. 
5 « CINGPAC-CINCPOA Trans 9645, 31st Army 
Operation Order 369-72 (Jun 44, Saipan). 

part, "It is a certainty that he [the 
enemy] will land in the Marianas Group 
either this month or next." HS 

The conclusion, then, is inescapable that 
the Japanese were simply not prepared for 
the American landings on Saipan when 
they came. In the short time intervening 
between the American capture of the Mar- 
shalls and the invasion of Saipan, the 
Japanese found it physically impossible 
with the means at hand to build up the 
defenses of the Marianas to a point where 
they might successfully resist a landing by 
U.S. forces. In their race against time the 
Japanese were further handicapped by the 
deadly efficiency of American submarines, 
which by now had extended the radius of 
their operations well to the west of the 
Marianas perimeter. Finally, U.S. forces 
achieved a measure of surprise by attacking 
Saipan before they were expected, thus 
adding the element of shock to the already 
preponderant strength they were able to 
bring to bear against this vital point in 
Japan's inner defensive perimeter. 

Prepared or not, the Japanese forces on 
Saipan were by no means feeble. The more 
than 30,000 troops on the island were 
backed by forty-eight tanks and an elabo- 
rate, if incomplete, network of artillery 
positions and reinforced by their determi- 
nation to die if necessary for the Emperor. 
They could be expected to put up a fierce 

5!i NTLF Rpt Marianas, Phase I, G-a Rpt, 1a 

Aug 44, P- 5- 



Softening the Target: 
Pre-D-Day Bombardment 

On 6 June, while the convoys carrying 
the attack troops headed westward for 
their staging bases in the Marshalls, Marc 
Mitscher's Task Force 58 weighed anchor 
and slipped out of Majuro for waters east 
of the Marianas. For this operation Admiral 
Mitscher had gathered together a total of 
seven carriers, eight light carriers, seven 
fast battleships, three heavy cruisers, ten 
light cruisers, and fifty -two destroyers. 1 
Their missions were to prevent Japanese 
aircraft from interfering with the capture 
of Saipan, Tinian, and Guam; to protect 
the expeditionary force and the troops 
ashore from attack by enemy surface ves- 
sels; and, commencing on D minus 3 (12 
June), to destroy aircraft and air facilities 
in the Marianas. Finally, on 13 June when 
it was presumed that Japanese aircraft op- 
erating from fields in the Marianas would 
be eliminated, Task Force 58 was directed 
to destroy all other types of Japanese de- 
fenses both by aerial bombardment and by 
ships' fire from its supporting vessels. 2 

This was to be the culmination of an 
accelerated program of aerial neutralization 
of the Marianas. Mitscher's fast carriers 


Anncx A, p, 28. 

2 Ibid., p, 22- 

-Jun 44, 

had raided the islands on 22-23 February, 
and a few bombs had been dropped in 
April on both Saipan and Guam by B-24's 
of the Seventh Air Force escorting Navy 
photographic planes over those islands, 3 
For almost three months Army heavy 
bombers and Navy and Marine Corps 
fighters and dive bombers had steadily 
pounded Truk, the western Carolines, the 
Palaus, and Marcus and Wake Islands. 
After the destructive carrier strike against 
Truk on 17 February, primary responsibil- 
ity for neutralizing that base as well as sis- 
ter islands in the Carolines fell to planes of 
the Seventh Air Force, stationed in the 
Marshalls, and the Thirteenth Air Force, 
based at Empress Augusta Bay, Bougain- 
ville Island, and later (early May) in the 
Admiralities. The neutralization plan had 
called for almost daily attacks, since Japa- 
nese runways could otherwise have been 
quickly repaired to accommodate replace- 
ment planes flown down through the chain 
of mandated islands. As the target day for 
Saipan approached, B-24 raids against 
Truk were stepped up sharply and other 
possible danger points in the Carolines 
were hit proportionately. 4 Meanwhile, late 
in May Mitscher's Task Group 58.6 had 
conducted a successful raid on Marcus and 
Wake Islands to eliminate the possibility of 

3 Sec Craven and Gate, AAF IV, pp. 684-87. 
1 Ibid., pp. 677-88. 



intervention from that direction in the 
forthcoming Marianas operation. 5 Thus, 
with all possible routes of enemy aerial in- 
terception from the south, cast, and north- 
east successfully interdicted, Task Force 58 
was assured a relatively free hand to deal 
with Japanese airpower based on the Mari- 
anas themselves. 

The original plan for the prcinvasion 
bombing of the Marianas called for the first 
Task Force 58 carrier strike to be launched 
at dawn of 12 June, three days before the 
scheduled landing. Because of unexpected- 
ly good weather conditions en route, the 
escorting destroyers were fueled more 
rapidly than had been anticipated, and the 
entire force arrived at points within fighter 
range of its targets earlier than planned. 
This bit of good fortune induced Admiral 
Mitscher to request permission to launch 
his first fighter sweep on the afternoon of 
1 1 June rather than wait until the follow- 
ing morning. His main reason was that all 
previous carrier attacks by Task Force 58 
had been launched at dawn and that an 
alteration in the pattern would surprise 
the enemy and be that much more effec- 
tive. Admiral Spruance approved, and at 
1300 on the nth the first planes took off 
from the carriers, which at that time were 
approximately 192 miles northeast of 
Guam and 225 miles southeast of Saipan 
and Tinian. The results were altogether 
gratifying. Of the 225 planes launched in 
this initial fighter sweep, only twelve were 
lost. By contrast, the enemy suffered heavi- 
ly. Estimates as to Japanese aircraft put 

out of operation either through destruction 
or serious damage ran from 147 to 215. 

Ashore on Saipan a Japanese soldier, 
member of the gth Tank Regiment, wrote 
of the strike in his diary : 

At a little after 1300, I was awakened by 
the air raid alarm and immediately led all 
men into the trench. Scores of enemy Grum- 
man fighters began strafing and bombing 
Aslito airfield and Garapan. For about two 
hours, the enemy planes ran amuck and fin- 
ally left leisurely amidst the unparallelledly 
inaccurate antiaircraft fire. All we could do 
was watch helplessly. At night we went to 
extinguish the mountain fires which had been 
caused by gun fire. They were finally brought 
completely under control. 7 

In spite of the magnitude of the attack, 
the Japanese command on Saipan appar- 
ently did not realize on the 11th that this 
was the prologue to a full-size invasion. At 
1600 on that date 43d Division headquar- 
ters ordered the construction of a new road 
between the Marpi Point and Aslito air- 
fields. The north^south highway already in 
use ran along the west coast adjacent to 
the ocean shore, and General Saito felt that 
in "the event of a battle occurring at the 
shore, there would be a great danger of 
the direction of the battle being hindered 
by an immediate interruption of com- 
munications." s The new road was to be 
inland from the coast line and follow the 
comparatively well-concealed foot of the 
mountains. Nothing could illustrate more 
graphically the Japanese failure to grasp 

5 Lt. (j.g.) A. O. Van Wyen and Lt. (j.g.) W. 
G. Land, Office, DCNO, Naval Air Operations in 
the Marianas: I 1-20 June 1944, pp. B-10-B-17, 
copy in Records and Research Sec, Hist Br G-3, 

a Ibid., pp. D-i-D-4, D-6; Coradr Fifth Fleet 
to COMINCH, Initial Rpt on Opn to Capture 
the Marianas, 13 Jul 44. 

7 CINCPAC-CINCPOA Trans 10238, Extracts 
From the Diary of Matsuya Tokuzo, gth Tank 

* CINCPAC-CINCPOA Trans 9378, 43d Divi- 
sion Operation Orders, pp. 27, 28, 29 (Saipan, 
Jim 44 ) . 



the fact that the 1 1 June bombing was not 
merely another hit-and-run strike but the 
beginning of an invasion. If it had sus- 
pected an immediate invasion, the Army 
command on Saipan would not have 
diverted to a long-range project when men 
and materiel that could and should have 
been devoted to emergency fortifications. 

Though there was a wide discrepancy 
in the estimates of damage inflicted on the 
Japanese during the attack of 1 1 June, 
there was no doubt that the enemy's power 
of aerial resistance in the Marianas had 
been considerably reduced. At no time 
thereafter were Japanese land-based air- 
craft more than a minor nuisance to 
American operations. According to Admiral 
Nimitz, "Control of the air had been 
effected by the original fighter sweep on 
1 1 June." 

Tor the next three days (12-14 June) 
all four of Mitscher's task groups flew 
scheduled strikes over Saipan, Tinian, 
Guam, Rota, and Pagan with the object 
of continuing the destruction of enemy 
aircraft, rendering airfields at least tempo- 
rarily useless, destroying coastal defense 
and antiaircraft batteries, and burning 
cane fields south of Mutcho Point on Sai- 
pan to prepare for forthcoming troop land- 
ings. In addition, last-minute photographic 
missions were flown over all three of the 
larger islands. During this period another 
fifty planes were reported destroyed with 
an additional sixty-six put out of operation. 
The task groups were less successful in 
bombing enemy airfields. Few runways on 
these or any other outlying bases were sur- 
faced with concrete, macadam, or steel 
strip since the comparatively light weight 
of Japanese aircraft made such expenditure 

of time and material unnecessary, and it 
proved almost impossible to render the 
earthen airfields permanently unserviceable 
by moderate bombing attacks, 10 

The effectiveness of preliminary aerial 
bombardment of coastal defense and anti- 
aircraft artillery is difficult to assess. Pilots 
reported direct hits on gun positions on all 
three islands, but the accuracy of these 
reports could not be precisely measured. 
The mere fact that enemy guns remained 
silent after a strike was no indication that 
they had been destroyed or even seriously 
damaged since the Japanese might have 
been holding their fire in order to save am- 
munition or avoid detection. Indeed, one 
dive bomber squadron leader after a run 
on Tinian admitted, "The odds of a dive 
bomber hitting a target the size of a gun 
are astronomical even under ideal condi- 
tions." He concluded that, on the basis of 
photographs and observations, shrapnel 
and blast resulting from the bombing 
caused the chief damage to enemy installa- 
tions, knocking out the control posts and 
damaging some of the guns. 11 

On 12 June Admiral Mitscher's carrier 
pilots came into an unexpected windfall 
in the form of two Japanese convoys trying 
to escape the area. One of these, composed 
of about twenty vessels and located about 
125 miles west of Pagan on a northerly 
course, was immediately bombed and 
strafed heavily. 12 Nine merchant ships, 
with a total tonnage of almost 30,000 tons, 
along with their escort vessels including 
one large torpedo boat, three submarine 
chasers, and a converted net tender, were 

» CTNCPAC-CINCPOA Opns in POA— Jun 44, 
Ann<*x A, p. 30. 

10 V;in Wyen and Land, Naval Air Operations 
in the Marianas, pp. C-2-C-r,. 

1 1 Ibid., p. G-5- 

44 PP- 30-31- 



sunk. 13 On the same day, other carrier 
planes hit two cargo vessels just off the 
northwest coast of Saipan, sinking one and 
damaging the other so badly that it had 
to be beached. Still another was sunk while 
being repaired in Tanapag Harbor. 14 On 
1 3 June a convoy fleeing south of the west 
coast of Guam was struck by planes of 
Rear Adm. Joseph G. Clark's Task Group 
58.1. One high-speed transport was defi- 
nitely sunk and other shipping was re- 
ported set on fire. 1 " 

Also on 1 3 June, while the carrier planes 
continued their bombing and strafing mis- 
sions against the islands, the fast battleships 
and certain designated destroyers were 
detached from escort and screening duties 
and assigned the mission of initiating naval 
shore bombardment of Saipan and Tinian 
and covering mine-sweeping operations. 
Seven fast battleships and eleven destroyers 
were detached and formed into a separate 
bombardment group under command of 

1S JANAC, Japanese Shipping Losses, pp. 12, 
60-61. Naval pilots mistook this torpedo boat, the 
Otori, for a destroyer and so reported it. (Van 
Wyen and Land, Naval Air Operations in the 
Marianas, p. C-27.) The error is understandable 
since the Otori, though less than half the size of a 
destroyer, resembled it somewhat in silhouette. 
Otherwise, American damage claims for this action 
erred on the side of modesty. Postwar studies in- 
dicated a total of fourteen ships sunk, whereas 
the official American Navy claim came only to 
ten. Van Wyen and Land, Naval Operations in 
the Marianas, p. C-27; JANAC, Japanese Ship- 
ping Losses, pp. 12, 60. 

11 JANAC, Japanese Shipping Losses, pp. 60- 
61; Japanese Studies in World War II, 97, A-GO 
Operations Log, pp. 13-14, OCMH. The Joint 
Army-Navy Assessment Committee claimed that 
one of these, the Keiyo Maru, was sunk. Japanese 
sources examined since the war indicate that it 
was only badly damaged and was subsequently 
beached on Saipan. 

lH JANAC, Japanese Shipping Losses, p. 12; 
Van Wyen and Land, Naval Air Operations in the 
Marianas, p. C-r 3 ; CINCPAC-CINCPOA Opns 
in POA — Jun 44, Annex A, p. 31. 

Vice Adm. William A. Lee, Jr. From 1 040 
until about 1725 they pounded the west 
coast of Saipan and Tinian. Meanwhile, 
ten fast mine sweepers probed the waters 
off the west coast of Saipan from distances 
of about six to two miles offshore. They 
found no moored contact or acoustics 
mines and received no fire from the beach. 
That night the battleships withdrew, but 
five destroyers remained in the area to de- 
liver harassing fire. 16 

The results of the first day's naval gun- 
fire were doubtful. At the close of the day's 
bombardment, headquarters of the 31st 
Army reported that although the city 
streets in Garapan and Charan Kanoa had 
been almost destroyed, personnel losses had 
been relatively slight. 17 In spite of na- 
val reports of considerable damage done 
to shore installations, General Holland 
Smith's naval gunfire officers remained 
skeptical. In their opinion, the effectiveness 
of the firing by these ships of Mitscher's 
task force had been limited because of 
severe handicaps. With one exception, the 
fast battleships had received no continuous 
training in shore bombardment as had most 
of the old battleships. This type of firing, 
which required slow, patient adjustments 
on specific targets, was quite different from 
that normally experienced in surface en- 
gagements and called for specialized train- 
ing. Also, air spotters off the fast battle- 
ships had neither experience nor training 
in locating ground targets. Finally, because 
none of the ships was allowed to move 
closer than 10,000 yards (five nautical 

44, Annex A, pp. 34-35, and Table XII, App. 1. 

17 CINCPAC-CINCPOA Item 9983-85, transla- 
tion of captured Japanese document, SONAE 
Staff Msg 101 1. 



Destruction at Charan Kanoa 

miles) from the shore for fear of mines, 
accurate fire against anything but large 
buildings and other such obvious targets 
was virtually impossible. 18 

Nevertheless, to the Japanese on the is- 
land the bombardment of the 13th, and 
especially that of the naval vessels, was a 
terrifying experience. One soldier described 
it thus: 

At 0500 there was a fierce enemy air at- 
tack. I have at last come to the place where I 
will die, I am pleased to think that I will die 
calmly in true samurai style. Naval gunfire 
supported this attack which was too terrible 
for words. I feel now like a full-fledged war- 
rior. Towards evening the firing died down 

18 TF 56 Rpt Forager, Incl G, Sec. 2, Rpt on 
Naval Gunfire Support, p. 10; NTLF Rpt Mari- 
anas, Phase I, Incl I, Sec. 3, Naval Gunfire Rpt, 

P- 5- 

but at night naval gunfire continued as 
before. About 1700 communications with bat- 
talion headquarters were cut off. lfi 

Another eyewitness, a Japanese naval 
officer, noted: "The shells began to fall 
closer and closer to the airfield. It was 
frightful. The workers were all rather de- 
pressed. 1 ' 20 The same officer reported that 
shortly after the naval shelling started he 
ordered his lookouts, his fire-fighting unit, 
and his workers to withdraw to caves in 
the hills. He himself remained behind with 
a junior officer and a "superior petty of- 
ficer." "On the veranda of the destroyed 

1!J CINCPAC-CINCPOA Trans 10051, Extracts 
From the Diary of an Unidentified Japanese 

30 CINCPAC-CINCPOA Translations and In- 
terrogations, Vol. 29, Trans B-1938, Diary of a 
Naval Officer, June-July 1944. 



workers' quarters," he notes, "we who had 
stayed behind bolstered our spirits with five 
bottles of beer." 21 

Early on the morning of 14 June, Rear 
Adm. Jesse B. Oldendorf arrived off the 
coast of Saipan with the two bombardment 
groups that would carry the main bur- 
den of naval gunfire support both before 
and during the seizure of the island. This 
force consisted of seven old battleships, 
eleven cruisers, twenty-six destroyers, and 
a few destroyer transports and fast mine 
sweepers. 22 The battleships had all been 
commissioned between 1915 and 1921. 23 
Four of them, California, Pennsylvania, 
Maryland, and Tennessee, were survivors 
of the attack on Pearl Harbor. 24 All had 
undergone the rigorous training program 
for shore bombardment set up by V Am- 
phibious Corps at Kahoolawe Island in the 
Hawaiian Islands. 

These ships were able to move into closer 
range than had the fast battleships in the 
previous day's bombardment. Mine sweep- 
ers had reported the area to the seaward of 
two miles from the shore line free of mines 
and were steadily moving in closer to the 
reef line. Better results were reported from 
this day's activities, and many installations 
were believed to have been directly hit, in 
spite of the facts that the time allowed for 

21 The consumption of five bottles of beer by 
three men is not as harmless as it would appear at 
first glance. Japanese beer is stronger than the 
standard Western brew and normally comes in 
bottles holding a little more than a quart. Also, 
Japanese are notoriously poor drinkers, with con- 
sumption and sobriety levels far below Western 
standards. This particular naval officer may have 
been an exception, for, as noted below, he resorted 
to stronger stuff on D Day. 

44, Annex A, Table IV, App. 1. 

3,1 Jane's Fighting Ships, 1944.-4.5 (New York, 
The Macmillan Company, 1947), pp. 436-46. 

2i Morison, Rising Sun in the Pacific, pp. 106- 
o8, 11 1-13, 1 17-19. 

deliberate pinpoint fire was too short and 
that air spotters again revealed their lack 
of training in distinguishing important land 

There is evidence that this preinvasion 
bombardment was especially effective a- 
gainst prepared gun positions of antiair- 
craft units, which were for the most part 
fixed. Two prisoners of war taken on 29 
June reported that their antiaircraft unit, 
the 1st Battery, 25th AAA Regiment, had 
been annihilated before D Day in the 
Magicienne Bay area. 2 '" 5 The Japanese naval 
officer quoted above noted in his diary, 
"Practically all our antiaircraft gun and 
machine gun positions were destroyed by 
bombing and shelling on the 13th, 14th, 
and 15th." 26 

In other respects, however, the American 
preliminary bombardment was far from 
perfect. A Japanese artillery instructor, as- 
signed to Saipan as an observer, managed 
to radio the following report on the effects 
of the shelling: 

Beach positions withstood four days of 
bombardment. Those observation posts and 
gun emplacements that were protected by 
splinter-proof shelters were able to withstand 
the bombardment. Dummy positions proved 
very effective. During bombardment, both 
day and night, movement to alternate po- 
sitions was very difficult. Communication 
lines were cut frequently, and the need for 
repairs and messengers was great. 27 

During this naval bombardment of the 
14th, two of the supporting ships were hit 
by fire from the shore. The destroyer 
Brajne, while bombarding Tinian, took a 
4.7-inch shell that caused three deaths and 

25 NTLF Rpt Marianas, Phase I, G-a Rpt, 12 
Aug 44. P- 86. 

36 CINCPAC-CINCPOA Trans B-1938. 

27 CINCPAC-CINCPOA Trans 15282, Tactical 
Lessons Learned on Saipan, by Captain Shima- 
mura, Combined Fleet, Aug 44. 



Dummy Searchlight Platform on East Coast of Saipan. False positions 
such as this effectively diverted some air and sea bombardment from genuine de- 
fense positions. 

numerous injuries. The battleship Cali- 
fornia was struck by a small caliber artil- 
lery shell; one man was killed, nine were 
wounded, and the ship's fire control sys- 
tem was damaged. 28 

Also on the 1 4th, three naval underwater 
demolition teams reconnoitered the landing 
beaches of Saipan as well as other parts of 
the shore line. Each team consisted of about 
sixteen officers and eighty men, all naval 
personnel except for one Army and one 
Marine liaison officer per team. The men 
were dispatched from destroyer transports 

28 TF 51 Opn Rpt Marianas, Incl A, p. 5; CO 
USS California, Rev Rpt of Participation in Sai- 
pan Opns, 7 Jul 44, p. 9. 

in small boats to the edge of the reef 
whence they swam close into the shore line 
in full daylight under the protection of 
ships' fire. No obstacles were reported and 
hence no demolition work was necessary. 
The teams performed their work under 
considerable fire from the beach, but even 
so only two men were killed and fifteen 
wounded — a low figure considering the in- 
herent danger of the operation and the fact 
that promised air support failed to materi- 
alize and ships' fire was generally too far 
inland to provide much protection. 29 

One result of the underwater demolition 
activities was to alert the Japanese on Sai- 

29 TF 51 Opn Rpt Marianas, Incl G, pp. 2-3. 



Japanese type 96 25-MM. Machink Gannon. U.S. marines repaired this damaged 
twin-mount weapon fitted with speed ring sights and promptly put it to use against 
the enemy. 

pan as to the probable place and time of 
the forthcoming landing. At 0800 a 31st 
Army message stated: "The enemy at about 
0730 was making reconnaissance of reef 
with small boats. It is judged that the en- 
emy will land here." Later in the day 
another message from the same headquar- 
ters reported; 

Since early this morning the enemy small 
vessels have been planting markers and 
searching for tank passages on the reef. Be- 
cause as far as one can see there are no trans- 
ports, the landing will have to be after 
tonight or dawn tomorrow. The enemy 
bombardment is being carried out on coastal 
areas in anticipation of a landing. so 

D-Day Bombardment and 
Ship-to-Shore Movement 

On the night of 14-15 June most of the 
support ships retired, leaving a handful to 
continue harassing fire along the coast line. 
Meanwhile the Western Landing Group, 
commanded by Admiral Hill and consist- 
ing mostly of transports and LST's carry- 
ing the 2d and 4th Marine Divisions, was 
slowly approaching the island from the 
east. As dark fell the marines could observe 
fires burning ashore and the glow of star 

: >o CINCPAC-CINGPOA Trans 9983-85, 31st 
Army Msg File, Msgs 1012, 10 15. 



shells fired by the naval ships left in the 
area. Shortly after 0500 the gigantic con- 
voy moved into the transport area off the 
west coast of Saipan. In the early dawn, 
Mount Tapotchau lay silhouetted in the 
cast. Streaks of fire from the armada of 
naval support ships colored the sky and the 
shore was blurred in a haze of smoke and 
dust. As the light improved, the town of 
Gharan Kanoa became visible. To the 
north lay Garapan, the capital city. In its 
harbor, called Tanapag, lay several Jap- 
anese ships, beached, half-sunken, and 

Naval bombardment commenced about 
0530. Heavy close support ships were or- 
dered not to approach closer than 1,500 
yards from the reef. Destroyers were per- 
mitted to move in as far as 1,000 yards. 
Two old battleships, two cruisers, and 
seven destroyers were assigned the duty of 
last-minute preparation of the landing 
beaches themselves. 31 At dawn these ships 
took station and shortly thereafter the two 
battleships commenced main battery fire at 
the beach defenses; less than an hour later 
the two cruisers opened up with their 12- 
inch guns. 

In spite of the apparent intensity of this 
barrage, the Japanese high command was 
not overly impressed— at least not officially. 
From 31st Army headquarters came the 
report: "They did not carry out large scale 
shelling and bombing against the positions 
on the landing beach just prior to landing. 
When they came to the landing, even 
though we received fierce bombing and 
shelling, our basic positions were complete- 
ly sound." 32 

44, Plate XV, App. A. 

32 CINCPAC-CINCPOA Trans 9983-85, 
Array Msg File, Msg 1037. 

But to other less exalted defenders of the 
island, the shelling appeared more formid- 
able. One member of the gth Tank Regi- 
ment observed fairly effective results from 
the shelling of the Magicienne Bay area. 
A naval supply warehouse was hit, 
causing a considerable number of casual- 
ties, and a nearby ammunition dump was 
set off. "There was no way," he reported, 
"of coping with the explosions. We could 
do nothing but wait for them to stop." 33 
Somewhere in the same vicinity, the Japa- 
nese naval officer mentioned above took 
to the bottle again to calm his nerves 
against the shock of the shelling. "I quietly 
opened the quart I brought along," he 
noted in his diary, "and took my first 
'shot' from it. There is something undescrib- 
able about a shot of liquor during a 
bombardment." 34 

At 0545 the word was passed throughout 
the American task forces that H Hour, the 
moment at which the first troops were sup- 
posed to land, would be 0830, as scheduled. 
Guns and winches were manned; boats 
were lowered into the water from the 

Shortly after 0700 the thirty-four LST's 
carrying the Marine assault battalions 
moved into position and dropped anchor 
about half a mile off the line of departure. 
The line, the starting point from which the 
assault landing craft would take off, was 
located 4,250 yards offshore. Bow doors 
swung open; ramps lowered, and hundreds 
of amphibian tractors and amphibian tanks 
crawled into the water and commenced to 
circle. In all, 719 of these craft would be 
employed in the operation. 

Astern of the assault landing ships lay 
twelve other LST's carrying light artillery, 

33 CINCPAC-CINCPOA Trans 10238. 

34 CINCPAC-CINCPOA Trans B-1938. 



most of which would be landed in 
DUKW's. Still further seaward of each 
division's beach lay two dock landing 
ships embarked with heavier landing craft 
(LCM's) that would take ashore the 
tanks and heavy artillery as soon as enough 
beachhead had been secured by the infan- 
try. 30 About 18,000 yards offshore the 
larger troop transports swung at anchor. 
Aboard were reserve troops, headquarters 
troops, shore party teams, heavy artillery, 
trucks, tractors, bulldozers, and sundry 
equipment and supplies. 

Meanwhile, north of the main transport 
area, marines of the 2d Regimental Combat 
Team (2d Marine Division) and the 1st 
Battalion, 29th Marines, were conducting 
a diversionary demonstration off the town 
of Garapan. Boats were lowered, troops 
embarked, standard waves were formed 
and went in as far as 5,000 yards off the 
beach. There they circled for ten minutes 
without receiving any fire and then re- 
turned to their mother ships. 

Off the main landing beaches, all ships' 
fire ceased at 0700 to allow a thirty-minute 
strike by carrier planes. Fifty fighters, 51 
scout bombers, and 54 torpedo bombers 
conducted an area bombing attack along 
the beaches with the primary aim of de- 
moralizing the enemy rather than knocking 
out particular installations. As soon as this 
strike lifted, the naval ships assigned to 
close support took up the course once 
again and continued to hit the beaches 
with heavy guns until the first wave of 
troops was only 1,000 yards from the 
shore line and with 5-inch guns until the 

troops had progressed to within 300 
yards. 36 

The line of departure was marked by 
four naval patrol craft (PC's), each an- 
chored and flying flags designating the 
number and color of the beaches opposite 
them. At 0750, H Hour was postponed to 
0840 because of a delay in launching the 
amphibian tractors. Small control craft es- 
corted the leading waves toward the line 
of departure. A few minutes after 0800 the 
control craft hoisted their Wave-i flags 
and twenty-four LCT gunboats crossed the 
line of departure, firing their automatic 
weapons as they went. About five minutes 
later wave flags were run down from the 
signal yardarms of the anchored patrol 
craft and the first wave of amphibian tanks 
and tractors crossed the line of departure. 
Following waves- three for the northern 
beaches and four for the southern — were 
spaced at intervals of from two to eight 
minutes. The run into the beach would 
take a few minutes less than half an hour 
at maximum LVT(A) speed of 4.5 knots. 
About 1 ,600 yards from shore the gunboats 
on the northern beaches stopped engines 
and lay to just short of the reef but kept 
up their fire as the first wave passed 
through them. On the southern beaches, 
where the reef was closer to the shore, the 
LCI(G)\s moved in to 400 yards and let 
loose salvos of 4.5-inch rockets in a last- 
minute saturation of the beach. As the lead- 
ing troops came within 300 yards of the 
shore, all naval gunfire ceased except in 
the area around Afetna Point, which lay 
between the two divisions' beaches. A last- 

,,)s LCM — landing craft mechanized — a shallow 
draft vessel, fifty feet long, equipped with a bow 
ramp and primarily designed to carry a tank or 
motor vehicles directly onto the beach. 

:i,i This account of D-Day preliminary bom- 
bardment is taken from the following sources: 
Annex A, pp. 41-47; TF 56 Rpt Forager, Incl 
G, Sec. 2, pp. 40-41; ad Marine Div SAR 
Forager, Phase I, Sec. V. 



Amphibian Tractors in Line Abreast Formation chum toward Saipan's 

minute strafing attack by seventy-two 
carrier-based planes commenced when the 
leading waves were 800 yards from the 
shore line, continued until the first troops 
were within 100 yards of the beaches, then 
shifted to 100 yards inland until the first 
landings were made. 

The formation of the assault waves dif- 
fered between the two Marine divisions. 
North of Afetna Point, the 2d Marine 
Division was landing with four battalion 
landing teams abreast. From north to south 
(left to right) the 6th Marines headed for 
Red Beaches 2 and 3 ; the 8th Marines for 
Green Beaches 1 and 2, immediately in 
front of the Charan Kanoa airstrip. South 
of Afetna Point, the 4th Marine Division 
proceeded toward Blue and Yellow Beaches 
with the 23d Regimental Combat Team on 
the left and the 35th on the right. 

In the 2d Marine Division's zone the 
first wave consisted of eight separated lines 
of six amphibian tractors, each in line- 
abreast formation. Between each line of six 

LVT's was echeloned one platoon of am- 
phibian tanks (LVT(A) (4)'s) mounting 
75-mm. howitzers. The succeeding three 
waves consisted of LVT's alone in line- 
abreast. The amphtracks (LVT's) were 
crewed by the Marine's 2d Amphibian 
Tractor Battalion and the Army's 715th 
Amphibian Tractor Battalion. The amphib- 
ian tanks, seventy in number, belonged to 
the Marine's 2d Armored Amphibian Bat- 
talion. m 

The 4th Marine Division's landing plan 
differed. The first wave consisted exclusive- 
ly of sixty-eight amphibian tanks, formed 
abreast and manned by the Army's 708th 
Amphibian Tank Battalion. Most of these 
were old style LVT(A)(i)'s with only a 
37-mm. gun on the bow, but sixteen of 
them were LVT(A) (4)'s carrying 75-mm. 
howitzers. Astern in four successive waves 

37 Army and Marine Corps nomenclature for 
these vehicles differed. In the Marine Corps they 
were designated "armored amphibians"' in the 
Army, "amphibian tanks." The latter usage will be 
followed in this volume. LVT's were sometimes 
called amtracs or amphtracks. 





IS June 1944 


i L««cmngs *s ticcurm 


Elrtetient m fttl 
a imo tooo moo 

F Tttnplt 

MAP 4 

came the assault troops boated in amphib- Company C, nth Amphibian Tractor Bat- 
ian tractors of the Marine i oth Amphibian talion) and the Army 773d Amphibian 
Tractor Battalion (less Company A, plus Tractor Battalion. 38 



From the line of departure to the reef 
the first waves moved in good order and 
met only moderate enemy gunfire. Once 
across the reef, however, the picture 
changed. All along the line the Japanese 
opened up with automatic weapons, anti- 
boat guns, and artillery and mortar bar- 
rages against the first wave. These in- 
creased in intensity as the second, third, 
and fourth waves climbed over the reef. 
In the 2d Marine Division's zone, three 
amphibian tanks and four tractors were 
knocked out of action between the reef 
and the bcach, aB Surf in the area ran as 
high as twelve to fifteen feet, too high for 
amphtracks to operate with any great de- 
gree of safety. Nevertheless, only two cap- 
sized as a result of the swells. About 98 
percent of all the tractors got ashore safely. 

Once across the reef, the wave forma- 
tion in the 2d Division area broke down 
completely. Because of their superior speed, 
many of the tractors commenced to over- 
take their supporting amphibian tanks and 
to compress them from echelon almost into 
column formation. Some tractors crossed 

;iK This account of the organization, of the ship- 
to-shore movement is derived from the following 
sources: TF r , i Opn Rpl Marianas, Incl C, pp. 
2-3. COMINCH P-007, Ch. IV, p. 6, Diagram- 
matic Chart Showing Organization of Boat Waves, 
Saipan, NTLF Rpt Marianas, Phase I, Incl E, 
G-3 Rpt, pp. 13-14. Ibid., Incl H, LVT Rpt, 
with the following inclosures: Incl A, Rpt of ad 
Amph Trac En, pp. 5-7; Incl B, Rpt of 5th Amph 
Trac Bn, pp. a-3 ; Incl C, Rpt of iolh Amph Trac 
Bn, pp. 1, 4; Incl E, Rpl of 715th Amph Trac Bn, 
pp. 1-3; Incl F, Rpt of 773d Amph Trac Bn, p. 
5 ; Incl G, Rpt of ad Armd Arnph Tk Bn, pp. 
r-a; and Incl H, Rpt of 708th Amph Tk Bn, pp. 
1-2. 2d Marine Div SAR, Phase I, Forager, Part 
I, Sec. V, p. 1. 4th Marine Div Opns Rpt Saipan, 
Annex G, pp. 1-9. 

39 NTLF Rpt Marianas, Phase I, Incl H, LVT 
Rpt; Incl A, Rpt of 2d Amph Trac Bn, p. 7; Incl 
E, Rpt of 715th Amph Trac Bn, p. 3; Incl G, Rpt 
of 2d Armd Amph Bn, p. 1. 

in front of the tanks thus masking their 
fire. 10 

Even more serious, the Navy guide boat 
led the leading waves off course. This has 
been variously attributed to compass error, 
a strong drift of current to the northward, 
and the fact that extremely heavy fire was 
coming at the boats on the right flank from 
the area around Afetna Point. Whatever 
the reason, the entire right flank of the 
leading waves veered to the left, thus caus- 
ing a northerly shift along the entire line 
and considerable crowding in the center. 
The 2d Battalion, 8th Marines, which was 
scheduled to land on Green Beach 2, went 
ashore instead on Green t , where it became 
badly intermixed with the 3d Battalion of 
the same regiment. The two assault bat- 
talions of the 6th Marines landed about 400 
yards north of their assigned beaches, Red 
a and 3, 41 

The first assault waves of the 8th Ma- 
rines landed on the Green Beaches at 
approximately 0843; the last were being 
landed by 0900. On the Red Beaches the 
first assault wave landed at 0840, the last 
at 0908. 

To the south, in the area of the 4th 
Marine Division, the ship-to-shore move- 
ment was proceeding somewhat more 
smoothly. Here, the sixty-eight amphibian 
tanks of the Army's 708th Amphibian 
Tank Battalion constituted the entire first 
wave. All commenced to fire 75-mm. howit- 
zers or 37-mm. guns about 400 yards from 
the shore after mounting the reef. From 
each beach came answering fire of all types 
including mortar, small arms, and artillery. 
Japanese artillery markers — small flags on 
bamboo sticks that were apparently a part 

4 " Ibid., Annex G, Rpt of 2d Armd Arnph Bn, 

p. 1. 

11 Hoffman, Saipan, p. 50. 



Congestion on 2D Marine Division Beach, the result of the right-flank 
battalion landing north of its assigned area. Note disabled LVT(A)(4) at 
upper right. 

of the enemy's prearranged fire plan and 
for unknown reasons had not been removed 
by underwater demolition teams on the 
previous day — were scattered along the 
reef. Of the sixty-eight tanks in the first 
wave all but three arrived safely. One 
burned, one was swamped on the reef, and 
one received a direct hit from an antitank 
weapon firing from the shore at about 
twenty-five yards range. 42 

42 ist Lt. Russell A. Gugeler, FA, 1st I and H 
Sv, Army Amphibian Tractor and Tank Battal- 
ions in the Battle of Saipan, if, June-g July 1944, 
20 Jan 45, pp. 6-7, MS in OGMH. These figures 
differ from those given in the official action report 
of the 708th Amphibian Tank Battalion, which 

Astern of the tanks came the amphibian 
tractors of the Marine ioth Amphibian 
Tractor Battalion and the Army 773d 
Amphibian Tractor Battalion in four 
waves, spaced from two to six minutes a- 
part. Of the 196 troop-carrying tractors, 
only two failed to land their cargo; one 
was hit by a shell on the reef and the other 

states that two of its tanks were overturned and 
three lost due to maintenance difficulties. The 
report mentions no direct hits. (NTLF Rpt Mari- 
anas, Phase I, Incl H to I.icl Y, p. 1 ). The former 
figures arc accepted as more accurate because the 
author of that account gives a detailed description 
of the LVT(A) casualties in the battalion as well 
as the names and actions of the personnel in- 



developed mechanical difficulties. Between 
0843 and 0907 all of the leading waves 
with about 8,000 marines embarked were 
ashore. 43 

Breakdown of the Landing Plan 

From the outset, two factors marred the 
smooth execution of the landing plan. The 
first was the wide gap that had developed 
between the right battalion (2d Battalion, 
8th Marines) of the 2d Marine Division 
and the left battalion (3d Battalion, 23d 
Marines) of the 4th Marine Division. The 
landing plan had provided for a gap be- 
tween the two divisions. Troops were not 
to land on Afetna Point itself because of 
the reasonable fear that the enemy would 
have placed his heaviest concentration of 
artillery there to guard the only channel 
through the reef to the pier at Gharan 
Kanoa. 44 However, the distance between 
the two divisions was more than double 
that envisaged because the 2d Battalion, 
8th Marines, landed north of its assigned 
beach. Almost three days would elapse be- 
fore firm contact between the divisions was 

Perhaps more serious was the breakdown 
of the scheme to employ amphibian tanks 
and tractors to carry the assault inland 
from the water's edge. The basic plan for 
the landing on Saipan prescribed a blitz 
assault, continuous from shipboard inland 
to the first high ground. In the 2d Marine 
Division's zone of action the four com- 

panies of amphibian tanks were to proceed 
inland about three hundred yards to the 
tractor control line, cover the debarkation 
of the assault troops from their tractors, 
and support their advance to the first ob- 
jective line, which lay about 1,500 yards 
inland. 45 Tractors of the first wave were to 
accompany the LVT(A)'s to the tractor 
control line and there debark their troops. 
The succeeding three waves were supposed 
to discharge their troops on the beach. 46 

In the zone of the 4th Marine Division, 

the initial wave- amphibian tanks — was 

to lead the next two waves — tractors — all 
the way to the first objective line, which 
was located on the first high ground a mile 
inland. There, the tanks would deploy 
and support the troops as they debarked 
and moved forward. The fourth and fifth 
waves were to be discharged at the beach 
and mop-up areas bypassed by the leading 
waves. 47 

None of these plans succeeded complete- 
ly, and for the most part the scheme of 
employing amphibian tanks as land tanks 
and amphibian tractors as overland troop- 
carrying vehicles must be marked off as a 
failure. The LVT(A)'s had neither the 
armor nor the armament to withstand the 
terrible pounding from enemy artillery and 
supporting weapons that could be expected 
during this phase of the assault. Moreover, 

48 NTLF Rpt Marianas, Phase I, Incl H, LVT 
Rpt; Incl C, Rpt of 1 Oth Amph Trac Bn, p. 4; 
Incl F, Rpt of 773d Amph Trac Bn, p. 5, 

44 Information from Maj. Carl Hoffman, 
USMG, who commanded the right flank company 
(Company G) of the 8th Marines, to the author, 
30 March 1950. 

4n NTLF Rpt Marianas, Phase I, Incl H, LVT 
Rpt; Incl G, Rpt of 2d Armd Amph Bn, p. 1. 
The tractor control line should not be confused 
with the transfer control line, which was On the 
seaward side of the reef and marked the point at 
which troops and supplies would be transferred 
from small craft (LCVP's) to amphibian tractors. 

40 NTLF Rpt Marianas, Phase I, Incl B, Rpt of 
5th Amph Trac Bn, p. 5. 

47 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RCT Rpt, p. 29; Annex J, 25th RCT Rpt, 
P- 3- 



the LVT(A) J s were underpowered and 
were stopped by sand, trenches, holes, 
trees, and other minor obstacles, most of 
which a land tank could have negotiated 
with ease. 48 Once the tractors were out of 
the water, their hulls were exposed and 
they became easy targets for enemy fire, 
which their armor was too light to resist. 
On shore they were clumsy and slow. It 
proved far healthier for troops to extricate 
themselves from these death traps as fast 
as possible and find shelter in whatever 
natural protection the terrain and the veg- 
etation offered. 

On the 2d Marine Division's beaches the 
situation rapidly became chaotic. Trees, 
trenches, and shell holes stopped some of 
the tanks of the ad Armored Amphibian 
Battalion before they could even cross the 
beach. Between the beach and the tractor 
control line, twenty-eight LVT(A)'s, more 
than one-third the total number, were dis- 
abled. Only a few points of ingress from 
the beach inland could be discovered, and 
while the amphibian tanks were maneuver- 
ing up and down trying to locate the points 
they became almost hopelessly intermixed 
with the tractors of the first and succeed- 
ing waves. Up to the tractor control line 
infantry troops were able to maintain close 
contact with the tanks, but the eighteen 
LVT(A)'s that went beyond that point got 
little infantry support. Tanks fired indis- 
criminately among troops and tractors and 
in general merely added to the confusion 
instead of aiding the battalions they were 
supposed to support. 49 

4M NTLF Rpt Marianas, Phase I, Incl II, LVT 
Rpt; Incl G, Rpt of qd Armd Amph Trac Bn, 
pp. 8-9; Incl H, Rpt of 708th Amph Tk Bn, pp. 


40 Ibid., Incl G, Rpt of Bd Armd Amph Trac 
Bn, pp. 1-2; 6th RCT 2d Marine Div SAR, 
Phase I, Forager, Incl A, p. 1, 

Congestion was particularly bad in the 
area of Green Beach 1 , where the two bat- 
talions of the 8th Marines were trying to 
land at the same time because the right 
flank of the first wave had veered to the 
left. These troops were all embarked in the 
amphibian tractors of the 7 1 5th Amphibian 
Tractor Battalion. Directly ahead of them 
was a heavily wooded bank constituting an 
almost impassable barrier for tanks and 
tractors alike. The marines forthwith aban- 
doned their tractors and took cover behind 
the embankment. Within five minutes after 
the first wave touched shore, the second 
wave arrived and landed a little to its right. 
By the time the third and fourth waves had 
landed, the men on foot were being 
squeezed between the tractors to their rear 
and the Japanese to the front. All together, 
only two tractors were able to get beyond 
the beach, one making its way as far as 
the radio tower 700 yards inland. Two 
days later, the driver wandered back to the 
beach but was too shell-shocked to be able 
to remember how he had got that far or 
what had happened when he got there. 5 " 

On the southern beaches, the 708th Am- 
phibian Tank Battalion fared a little better. 
Of the sixty-eight tanks in the first wave, 
about half reached Pina Susu ridge by ten 
o'clock. 51 Contrary to expectations, prog- 
ress through the town of Charan Kanoa 
was fairly easy, and by 09 1 5 thirteen tanks 
of Company B, assigned to this sector, had 
arrived at the objective. South of them 
the going was more difficult. The tanks 

no Gugeler, Army Amph Trac and Tk Bns at 
Saipan, p. 22. 

51 Ibid, 3 p. 10. Gugclcr states that thirty-three 
tanks reached the O— 1 line, most of them by 
1000. According to the action report of the bat- 
talion itself, forty tanks were on the O-i line by 
1000. NTLF Rpt Marianas, Phase I, Incl H, LVT 
Rpt; Incl H, Rpt of 708th Amph Tk Bn, p. 3. 



that landed below Charan Kanoa had to 
fight swampy ground, tank trenches, heavy 
artillery, a burning gas dump, and a steep 
railroad embankment before attaining the 
ridge. Nineteen made it. Those on the ex- 
treme right had the most difficulty. Swing- 
ing south, three tanks reached the tip of 
Agingan Point about a thousand yards 
south of the lowest landing beach. They 
braved it out against fairly heavy Japanese 
mortar and artillery fire, but when Ameri- 
can naval shells commenced to drop in the 
area they discreetly withdrew/' 3 Agingan 
Point would have to wait another day 
for capture. 

Meanwhile, back on the beaches most 
of the troops were deserting their amphib- 
ian tractors for the dubious safety of 
traveling on foot and belly. The 23d Ma- 
rines, which landed in and just south of 
Charan Kanoa, made fair progress inland. 
Fifteen LVT's of the second wave were able 
to carry their troops through the town it- 
self and on to the ridge behind. On the 
beach immediately below, thirty-three trac- 
tors got as far as a railroad embank- 
ment about 400 yards inland before the 
infantry commanders ordered their troops 
to debark. sS 

On the southern (Yellow) beaches, 
where the fighting was fiercest, no such 
progress was made. One company of the 
773d Amphibian Tractor Battalion got as 
far as a railroad spur about 700 yards in- 
land, but the rest unloaded their troops as 
rapidly as possible and shoved off back to 
the transfer line beyond the reef. In spite of 
the hasty withdrawal of most of the first 
group of LVT's, the shore line soon was 

thick with tractors as succeeding waves tel- 
escoped onto the beach. It was small 
wonder. After a full hour's fighting the 1st 
Battalion, 25th Marines, had succeeded in 
pushing just twelve yards in from the 
beach, and the ad Battalion's progress was 
only a little better. K4 The tractor plan 
failed most signally here. The infantryman 
was on his own. 

Expanding the Beachead 
Action of the ad Marine Division 

On the Red Beaches to the north, the 
two assault battalions of the 6th Marines 
met fierce enemy fire immediately upon 
landing. The failure of most of the amphib- 
ian tanks and tractors to proceed any con- 
siderable distance inland and the rain of 
enemy shells caused unexpected congestion 
and confusion on the beaches. 55 The com- 
mand posts of both battalions received 
direct hits that seriously injured battafon 
commanders and most of their staffs. The 
regimental commander, Col. James P. 
Riscly, USMC, came ashore at 1000 and 
established his command post practically 
at the water's edge. Forty minutes later the 
regimental reserve (1st Battalion, 6th Ma- 
rines) commenced to land and prepare to 
support the assault elements. 

By 1 1 05 the front line had advanced only 
400 yards inland. The 3d Battalion, on the 
right, was suffering especially heavy casual- 

52 Gugcler, Army Amph Trac and Tk Bns at 
Saipan, pp. 10-1 I. 

53 NTLF Rpt Marianas, Phase I, Incl H, LVT 
Rpt; Incl C, Rpt of 10th Amph Trac Bn, p. 4. 

54 4th Marine Div Opns Rpt Saipan, Annex J, 
25th ROT Rpt, p. 3, 

55 The following account of D-Day activity in 
the 2d Marine Division's zone of action is derived 
from: 2d Marine Div SAR, Phase I, Forager, 
Sec. V, pp. 1-5; 6th RCT ad Marine Div SAR, 
Phase I, Forager, Incl A, pp. 1-2; 1st Bn 20th 
Marine SAR Saipan, Incl A, p. i ; Hoffman, 
Saipan, pp. 51-55. 



Early Wave on the Beach, protected by embankment. Black rods in foreground 
are portable radio antennas. Note disabled amtracks. 

ties, and the ist Battalion was therefore 
ordered to pass through the 3d. Weak 
points began to appear all along the line. 
More serious, a dangerous gap developed 
between the right flank of the 6th Marines 
and the left flank of the 8th Marines when 
the 6th Marines landed some 400 yards 
north of its assigned beaches. In spite of 
the fact that Companies K and L of the 
6th Regiment were ordered to establish 
physical contact, a gap of 300 yards still 
existed, although covered by fire. 

Around noon the 6th Regiment's casual- 
ties had mounted to an estimated 35 per- 
cent. An hour later three Japanese tanks 
counterattacked in the area in front of the 

command post. Two of the tanks were 
knocked out by bazookas before penetrat- 
ing the front line, but the third managed 
to push through to within seventy-five 
yards of the command post before being 
disabled by a bazooka rocket fired from the 
post itself. During the morning and early 
afternoon the regiment had no supporting 
weapons ashore other than those carried as 
organizational — bazookas, antitank gren- 
ades, and 37-mm. guns. Meanwhile, how- 
ever, some tanks had commenced to land. 
Shortly after 0900 a pilot tank was 
disembarked at the reef's edge to explore 
the best passage through the reef. By 1020 
it had searched out a path to Green Beach 



i, although it was under heavy fire all the 
way. Once on the beach, enemy fire forced 
the crew to abandon the tank, but a reef 
route was marked and by 1300 the first of 
the 2d Division's tanks had landed on 
Green 1 and moved northward to support 
the 6th Marines. By midafternoon those 
tanks assigned to the 8th Marines had 
successfully landed and were in operation. 
Southward, on the Green Beaches, the 
chief problems facing the 8th Marines were 
confusion and congestion. The right flank 
battalion (the 2d) had landed from 700 to 
1,000 yards north of its assigned beaches. 
As a result, the two assault battalions of 
the 8th Marines and part of one battalion 
of the 6th Marines all found themselves in 
the same beach area. To add to the con- 
fusion, the commanders of the two assault 
battalions of the 8th Marines were both 
wounded early in the action and had to 
be evacuated. The 2d Battalion, 8th Ma- 
rines, had the most difficult job — to attack 
south along the beach toward Afetna Point. 
The object was not only to remove the 
menace of antiboat weapons located in 
'.hat area but also to secure the single reef 
channel off Charan Kanoa and permit the 
early entry of tank-carrying landing craft, 
which could not negotiate the reef itself. 
According to their prescribed scheme of 
maneuver, one company (Company G) 
moved south along the beach and the other 
two (E and F) fanned out to the south- 
east. Company G was heavily armed with 
shotguns in addition to its normal weapons. 
These short-range guns with wide disper- 
sion patterns were allotted to Company G 
chiefly as insurance against its firing into 
the lines of the 4th Marine Division toward 
which it was advancing. Progress was slow 
— the beach itself was thickly covered with 

pillboxes, and enemy riflemen situated east 
of the Charan Kanoa airstrip made the 
most of the flat, open terrain to harass the 
company's left flank as it inched south- 
ward. At 0950 the i st Battalion, in regi- 
mental reserve, was ordered to land. 
Company B was attached to the 2d Bat- 
talion to support Company G's attack 
toward Afetna Point. Companies A and C 
were committed between the 2d and 3d 

Later in the afternoon, the 1st Battalion, 
29th Marines, was landed and also at- 
tached to the 8th Marines on the Green 
Beaches. Because of a shortage of amphib- 
ian tractors, the battalion was unable to 
boat up properly and hence landed in 
considerable disorder. Company B, 29th 
Marines, was ordered to proceed at once 
to fill a gap between Companies E and G 
of the 8th Marines. However, its knowledge 
of the terrain was inaccurate and it was 
furnished no guides, so the company ended 
up about 600 yards north of its assigned 
position. The gap between the two right 
companies of the 2d Battalion, 8th Ma- 
rines, still remained unfilled, and Company 
A, 29th Marines, was ordered at 1730 to 
take that position. The approach of dark- 
ness and a heavy barrage of enemy 
artillery pinned these troops down before 
they could arrive at their destination. r,<1 

Meanwhile, the 2d Marines, which had 
participated in the demonstration off Tana- 
pag, was beginning to come ashore on Red 
Beach 2. By 1800 the 3d Battalion had 
landed and was attached to the 6th Ma- 
rines, taking station before nightfall on the 
division left flank. By nightfall one com- 
pany from the 2d Battalion, 2d Marines, 

1 st Bn 29th Marines SAR Saipan, Iricl A, 

p. 1. 



Marines Digging Foxholes, checking weapons, and keeping a steady watch. 

in addition to the regimental commander 
and the advance echelon of his command 
post, was also ashore. The rest of the regi- 
ment was ordered to return to the area of 
the control vessel on the line of departure 
and remained boated throughout the night, 
to be landed by midmorning of the follow- 
ing day. 57 

By noon of D Day Red Beach 3 was 
sufficiently clear to permit shore parties to 
land. The first team came ashore at 1300, 
and supplies began to flow over the two 
central beaches. Two more shore parties 
landed before the end of the day. Late in 
the day the two 75-mm. pack howitzer bat- 
talions of the 10th Marines were ashore 

and in position to support the infantry. 
The 1 st Battalion landed by 1403 and 
supported the 6th Marines; the 2d Bat- 
talion was ashore at 1730 and in position 
to support the 8th Marines. Before dark 
the 2d Marine Division's commander, Gen- 
eral Watson, had established his command 
post on Red Beach 2. By this time the 
division was digging in for the night and 
consolidating its positions against coun- 
terattack. Amphibian tanks and tractors 
had set up a defensive net against pos- 
sible countcramphibious attacks from the 
sea. Division casualties were estimated to 
amount to 1,575—238 killed, 1,022 
wounded, and 3 i 5 missing in action/' 8 

■" ad Marines 2d Marine Div SAR, Phase I, 

FORAGKRj p. 2, 

58 3d Marine Div SAR, Phase I, Forager, Sec. 

V, P. 5- 



lilUI I 

D-Day Command Post, a captured Japanese gun position in the Red Beach area. 

Action of the 4th Marine Division 

South of Afctna Point, the 4th Marine 
Division was having it own share of prob- 
lems. 50 To be sure, opposition in the town 
of Charan Kanoa was comparatively light. 
Japanese riflemen sniped away as troops 
and tractors moved through the rubble of 
the town, but they caused small damage. 
The 3d Battalion, 23d Marines, reached the 
first objective line with phenomenal speed. 
On its right flank, however, the 2d Bat- 

r ' u This account of operations in the 4th Marine 
Division's none is derived from the following 
sources: 4th Marine Div Opns Rpt Saipan, pp. 
13-15; Annex C, Opns, pp. 9-12; Annex H, 23d 
RCT Rpt, pp. 30-35; Annex I, 24th RCT Rpt, 
p. iJi; Annex J, 25th RCT Rpt, pp. 3-4; Annex 
K, Rpt of 4th Tk Bn, p. 3; Hoffman, Saipan, pp. 

talion of the same regiment was not so 
lucky. Troops debarked from their tractors 
unevenly and there was no semblance of 
a continuous line. Fighting degenerated 
into a series of small unit actions, and not 
until midafternoon was tactical control re- 
gained by the battalion commander. The 
farther south, the worse the situation 
became. After a full hour, the 25th Marines 
had penetrated only twelve yards in from 
the beach. From its right flank the 1st 
Battalion caught the heaviest load of fire 
from pillboxes and mortars on Agingan 
Point. Amphibian tanks, bombs, and naval 
shells were unable to abate this nuisance 
for the remainder of the day, and at night- 
fall the 1st Battalion, 25th Marines, had 
to dig in with its right flank exposed. 



Tanks of the 4th Marine Division began 
to come ashore about two hours after the 
initial landing. Their progress from ship to 
shore was seriously impeded since the chan- 
nel off Afetna Point was still interdicted 
by enemy fire, thus making it necessary in 
most cases for tanks to debark from land- 
ing craft at the reef and attempt to 
negotiate the lagoon under their own 
power. Mounting seas during the afternoon 
increased the hazards of the trip. Of the 
sixty tanks of the 4th Tank Battalion, 
twenty-one failed to reach the 4th Marine 
Division's beaches. One sank with the 
LCM on which it was boarded, another 
settled into a pothole off the reef, others 
were unable to locate landing craft to take 
them ashore or had their wiring systems 
fouled en route. Finally, six medium tanks 
were misdirected to Green Beach 2, a 2d 
Division beach. Of these, five were immo- 
bilized in deep water inside the lagoon and 
the sixth was appropriated by the 2d 
Division and failed to reach its parent or- 
ganization until several days later. The 
only decisive tank action on the southern 
beaches occurred on the extreme right 
flank. There, the 1st Platoon of Company 
A helped to break up a counterattack that, 
if successful, would have driven the ist 
Battalion, 25th Marines, back into the 
sea. 60 

In the matter of artillery, the 4th Marine 
Division was more fortunate than the 2d 
Division. Whereas the latter got only two 
battalions of 75-mm. pack howitzers ashore 
on D Day, the entire 14th Regiment, con- 
sisting of two 75-mm. pack howitzer 

battalions and three 105-mm. howitzer 
battalions, was landed by 1630 to sup- 
port the 4th Marine Division. ' 1 Shortly 
thereafter the reserve regiment, the 24th 
Marines, landed and proceeded to an area 
about 800 yards south of Charan Kanoa/' 2 

At 1930 General Schmidt, commander 
of the 4th Division, came ashore. His com- 
mand post was a scries of foxholes about 
fifty yards from the beach and very poorly 
protected from enemy light artillery, which 
was firing from the high ground about 
1,500 yards away. General Schmidt later 
recalled, "Needless to say the command 
post during that time did not function very 
well. It was the hottest spot I was in 
during the war, not even excepting Iwo 
Jima." 63 

By the time the division commander had 
landed, the division's left flank had been 
pulled back to conform to the configura- 
tion of the remainder of the line. The 2d 
and 3d Battalions, 23d Marines, were both 
ordered to withdraw to positions roughly 
800 yards west of the first objective line 
on the reverse slope of Fina Susu ridge. 
The movement was executed under cover 
of darkness — a difficult operation but one 
carried out successfully and without alert- 
ing the enemy. After the withdrawal was 
completed, the ist Battalion relieved the 
3d, and the latter assembled in what was 

60 This account of 4th Marine Division tank 
action on D Day is derived from: 4th Marine Div 
Opns Rpt Saipan, Annex K, Rpt of 4th Tk Bn, 
pp. 2-3; Hoffman, Saipan, pp. 143-48. 

fil 4th Marine Div Opns Rpt Saipan, p. 14. At 
the time of the Marianas Campaign, the Marine 
Corps artillery regiment was ordinarily composed 
of only four battalions, two of 75-mm. pack howit- 
zers and two of 105-mm. howitzers. At Saipan, V 
Amphibious Corps attached an extra 105-mm. 
battalion to the 4th Marine Division and a 155- 
mm. howitzer battalion to the 2d Division. 

" 2 4th Marine Div Opns Rpt Saipan, Annex 1, 
24th RCT Rpt, p. 18. 

B;t Comments by General Schmidt on draft copy 
of this volume, OCMH. 



euphemistically called a "rear area" to 
protect the left flank. 04 

Summary of the Situation at Nightfall 

By darkness of the first day it could be 
concluded that the landing was a success, 
even though only about two thirds of the 
area within the first objective line was 
under the marines' control. In spite of the 
failure of the initial plan to carry some of 
the assault waves 1,500 yards inland 
aboard amphibian tractors, the troops 
had established a beachhead approximately 
10,000 yards in length and over 1,000 
yards in depth in most places. Two divi- 
sions were ashore with almost all their 
reserves. Seven battalions of artillery had 
landed, as had most of the two tank bat- 
talions. Both division command posts were 
ashore by the time the troops had dug in 
for the night. The most serious weaknesses 
in the Marine position were on its flanks. 
Afetna Point, between the two divisions, 
was still in enemy hands. So was Agingan 
Point on the right flank, and the 6th Ma- 
rines' hold on the extreme left was 

As for the Japanese, they had exacted 
a heavy toll — how heavy cannot be accur- 
ately stated because of the inadequacy of 
casualty figures for D Day. The American 
landings had been made against what the 
enemy considered his strongest point and 
at a time when his garrison there was four 
battalions ovefstrength. He had registered 
the landing area, using flags on the reef 
for registration markers, and as successive 
waves landed artillery and mortar fire in- 

creased in intensity. The Japanese had 
massed at least sixteen 105-mm. howitzers 
and thirty 75-mm. field pieces on the first 
high ground and the reverse slope thereof 
about 1.5 miles southeast of Charan 
Kanoa. Directly east of the airstrip they 
had emplaced a 150-mm. howitzer battery 
of four weapons with a similar battery 
south of it. All of these weapons were well 
sited, and they were responsible for a tre- 
mendous amount of fire on the landing 
beaches. fiS 

Although the enemy realized that the 
diversionary maneuver off Tanapag was a 
ruse, he did retain one infantry regiment 
(the 135th) in that area instead of com- 
mitting it, as was intended, to the south 
of Garapan. At no time on D Day did the 
Japanese employ infantry in any great 
strength. They relied almost entirely on 
artillery, heavy weapons, and scattered 
tank attacks. 

In the opinion of Holland Smith's op- 
erations officer, the "most critical stage of 
the battle for Saipan was the fight for the 
beachhead: for the security of the landing 
beaches, for sufficient area into which 
troops and heavy equipment could be 
brought, and for the ability to render 
logistical support to those forces once 
landed." (i,! This, to be sure, could be said 
of any amphibious landing where strong 
opposition is encountered. On Saipan, it 
was six days before the beachhead could 
be considered completely secured, but it 
was the first day's action that was crucial. 
The most critical stage of "the most critical 
stage" was past. 

a ' 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RCT Rpt, p. 34. 

fi5 TF 56 Rpt Forager, Incl D, G-2 Rpt, pp. 

6 « Ibid., G-;j Rpt, p. 5. 


Capture of Aslito Airfield 

Night of 15-16 June 

Nightfall brought little hope of respite 
to the battle-weary marines on Saipan 
as they dug in on their narrow strip of 
beachhead with the Philippine Sea at 
their backs and a vengeful and still potent 
enemy lurking in the dark ahead. All had 
been alerted to the strong possibility of 
a night counterattack. Few doubted that 
it would come — the only questions being 
where, when, and in what force. 

In fact, by midafternoon of the 1 5 th the 
Japanese high command on Saipan had 
already issued orders to drive the Ameri- 
cans back into the sea before daylight next 
day. To Tokyo, 31st Army radioed opti- 
mistically, "The Army this evening will 
make a night attack with all its forces and 
expects to annihilate the enemy at one 
swoop." 1 To the troops, the order went 
out, "Each unit will consolidate strategic- 
ally important points and will carry out 
counterattacks with reserve forces and 
tanks against the enemy landing units and 
will demolish the enemy during the night 
at the water's edge," 2 

First to feel the effects of these measures 
was the 6th Marines, 2d Division, which 

held the left flank of the beachhead. 3 
About 2000, a large force of Japanese in- 
fantry, supported by tanks, bore down 
from the north along the coastal road. 4 
With flags flying, swords waving, and a 
bugle sounding the Japanese fell upon the 
marines' outposts. Unhappily, the 2d Ma- 
rine Division had been able to land none 
of its 105-mm, howitzer battalions during 
the day so the regiment under attack had 
only one battalion of 75-mm. pack howit- 
zers to support it. However, naval star 
shells fired from American destroyers lying 
close off the coast silhouetted the attackers 
as they approached, and the first attack 
was stopped by the withering fire of ma- 
chine guns and rifles, assisted by naval 
5-inch guns. 

A second, though smaller counterattack 
developed in the same area around 0300 
on the 1 6th. It, too, failed to penetrate the 
marines' lines. Finally, just before daylight 
another organized force of infantry and 
tanks rolled down the road from Garapan, 
Again, the Japanese were repulsed, this 
time with the help of five American me- 

9983, p. 5- 

2 NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, p. g. 

3 The account of the night's action in the 2d 
Marine Division zone is taken from; 2d Marine 
Div SAR, Phase I, Forager, Sec. VI, p. 1; 6th 
Marines SAR Saipan, pp. 3-4; Hoffman, Saipan, 
pp. 71-72. 

4 The exact timing of the beginning of the main 
counterthrust is not precisely known since the re- 
ports vary. The time stated here is derived from 
the 6th Marines SAR Saipan, page 3. 



dium tanks. By dawn the full measure of 
the enemy's failure was revealed, About 
700 Japanese lay dead just to the north of 
the 6th Marines flank. 

In the zone of the 4th Marine Division, 
enemy countermeasures on the night of 
15-16 June were less well organized and 
less powerful. Also, the 4th Division had 
all three of its 1 05-mm. howitzer battalions 
ashore by nightfall and was in a better 
position to resist. On the southern beaches, 
small groups of enemy soldiers, one 
shielded by a spearhead of civilians, hit 
once at 0330 and again an hour later. Both 
thrusts failed, with much of the credit for 
the successful defense going to a battalion 
of 1 05-mm. howitzers. 

The most vulnerable spot in the 4th Ma- 
rine Division's zone of action, of course, 
lay on the exposed left flank, where the 
23d Marines had not yet tied in with the 
2d Division to the north. All through the 
night Japanese artillery fire swept the 
beaches in this area from one end to the 
other. From dusk to dawn small groups of 
the enemy managed to filter through front- 
line units only to be wiped out in the rear 
areas by either infantry or shore party 
personnel. Among the latter were the 3 1 1 th 
Port Company and the 539th Port Com- 
pany. These were attached to the 4th 
Marine Division and were the first Army 
units to be put ashore on Saipan. Finally, 
at 0530, about 200 Japanese launched an 
organized attack. Through the gap it came, 
apparently aimed at the pier at Charan 
Kanoa. It too was stopped. Only a 
few individual enemy soldiers reached the 

beaches, where they were disposed of by 
members of the shore parties. 

One important factor that contributed 
to the marines' success in warding off 
these early morning counterattacks was the 
bright illumination provided by the Navy. 
The battleship California, assisted by two 
destroyers, cruised off the west coast of 
Saipan all night firing star shells to light 
up danger spots from which surprise at- 
tacks might be launched. That they were 
highly successful was later confirmed by 
31st Army headquarters itself, "The enemy 
is under cover of warships nearby the 
coast; as soon as the night attack units go 
forward, the enemy points out targets by 
using the large star shells which practically 
turn night into day. Thus the maneuvering 
of units is extremely difficult." e 

In spite of precarious holds on both the 
extreme flanks and the gap in the middle 
between the two divisions, the marines 
therefore succeeded in maintaining their 
positions and thwarting all major efforts to 
drive them back into the sea. Those few 
Japanese who managed to infiltrate behind 
the lines were wiped out without causing 
any considerable damage. The enemy plan 
of maneuver had relied in the main on re- 
pelling the American assault troops at the 
beach by counterattacks with artillery and 
tanks in support. As dawn broke on the 
morning of 16th June, the miscarriage of 
the Japanese first basic defense plan was 
more than evident. 

Consolidating the Beachhead 
16 June 

5 The account of the night's action in the 4th 
Marine Division zone is taken from; 4th Marine 
Div Opns Rpt Saipan, p, 16; 23d Marines SAR 
Saipan, p. 35; 25th Marines Final Rpt Saipan 
Opn, pp. 16-17; Hoffman, Saipan, pp. 73-74. 

Daylight brought to the grateful marines 
hugging the beaches a respite at least from 

* 31st Army Msg File, Msg 1038. 



the fearful dread of night counterattacks 
and infiltration. But immediate and press- 
ing duties lay ahead. No more than a half 
of the designated beachhead (west of the 
O-i line) was under their control. (See 
| Map ^.)| Afetna Point had not been sc- 
cured, which meant that a gap of about 
800 yards lay between the two divisions. 
The tip of Agingan Point, the southwest 
extremity of the island, still remained in 
enemy hands. Finally, an unknown num- 
ber of Japanese could be presumed to be 
still lurking behind the lines, ready to am- 
bush the unwary and harass the attacking 
troops from the rear. 

On the left (north) flank, the 6th Ma- 
rines held fast and consolidated the posi- 
tions won the day before. South of the 6th, 
the 8th Marines made rapid progress in its 
zone of action. Afetna Point offered little 
resistance, and the few Japanese left there 
after the previous night's counterattack 
were quickly mopped up. By 0950 the right 
flank company of the 2d Marine Division 
had reached Charan Kanoa pier and 
about two hours later established contact 
with the left flank of the 4th Marine 
Division. 7 

The heaviest fighting of the day took 
place in the zone of the 4th Marine Di- 
vision, especially on its right flank. Orders 
called for the capture of all ground lying 
west of the O-i line along Fina Susu ridge 
by nightfall, but the assault was held up 
until 1230 while lines were rearranged. On 
the division right, the 25th Marines en- 
countered considerable opposition from 
machine guns, mountain guns, and the 
antiaircraft weapons guarding the western 
approaches to Aslito field. By the end of 

the day's fighting the 25th had overrun 
Agingan Point and accounted for five ma- 
chine guns, two mountain guns, and ap- 
proximately sixty Japanese combatants. 
Meanwhile, the left and center regiments, 
the 23d and 24th, moved abreast of the 
25th Marines and by 1730, when the fight- 
ing was called off, the lines of the 4th Ma- 
rine Division rested generally along the 
Fina Susu ridge line. 8 

On the same day, to the north of the 
main area of fighting, additional elements 
of infantry and artillery were being landed 
on the beaches controlled by the 2d Marine 
Division. By 1000 of the 16th those men of 
the 2d Battalion, 2d Marines, that had not 
come ashore on D Day were landed and 
took positions on the division left. 8 Around 
1600 the 1 st Battalion, 2d Marines, which 
had originally been scheduled to invade 
Magicienne Bay, 10 was landed, minus its 
heavy weapons, on the 2d Marine Divi- 
sion's beaches. The heavy weapons were 
subsequently dropped by parachute from 
carrier torpedo planes, but because the 
planes flew at a low altitude the equipment 
was almost completely destroyed. 11 

At the same time that the remaining in- 
fantry elements of the 2d Marine Division 
were being dispatched shoreward, the two 
105-mm. battalions of the 10th Marines 
were also going into position in the area. 12 
About 1600 the 4th Battalion landed just 
north of Afetna Point and set up its bat- 
teries to support the 8th Marines, while an 

7 8th Marines SAR Foragek, 20 Jul 44, pp, i-<?; 
6th Marines SAR Saipan, 18 Jul 44, pp. 4-5; 
Hoffman, Saipan, pp. 79-80. 

8 4th Marine Div Opns Rpt Saipan, Annex H, 
33d RCT Rpt, pp. 35-36; Annex I, 24th RCT 
Rpt, p. 18; Annex J, 25th RCT Rpt, p. 17; Hoff- 
man, Saipan, pp. 81-83. 

* 2d Marine Div SAR, Phase I, Forager, p. 2. 

10 See above, pp. 40-41. 

11 Hoffman, Saipan, pp. 80-81. 

1 s The two 75-mm. pack howitzer battalions had 
landed on D Day. 



hour later the 3d Battalion came ashore 
on Red Beach 3 behind the 6th Marines. 1S 
At 1515 on the 16th, General Harper, 
USA, commanding the XXIV Corps Artil- 
lery, left the flagship Cambria and an hour 
later arrived on Blue Beach 2 just south 
of Charan Kanoa. There, he set up his 
command post about a hundred yards in- 
land from the southern edge of Blue Beach 
2, and before dark advance parties of the 
149th and 420th Field Artillery Groups, 
the 225th and 531st Field Artillery Bat- 
talions, and elements of his staff reported 
to him there. No corps artillery equipment 
came ashore on 16 June, and the advance 
elements spent an uneasy night dug-in 
in a partially destroyed enemy gasoline 

Night of 16-17 June 

General Saito's failure to "drive the 
enemy back into the sea" the first night 
after the landing did not discourage him 
from making a second try. During the 
afternoon of the 16th he ordered the 136th 
Infantry Regiment and the gth Tank Regi- 
ment to launch a co-ordinated attack at 
1700 toward the radio station that now 
lay behind the lines of the Gth Marines. 
Another, through un-co-ordinated, attack 
was to be carried out by the Yokosuka 
Special Naval Landing Force from the di- 
rection of Garapan. 15 

The scheduled hour came and passed, 
but the units assigned to the task were 
apparently too disorganized to carry it out 

13 ioth Marines SAR, 22 Jul 44, pp. 2-3. 

14 XXIV Corps Arty Final Rpt on Fokagkk 
Opn, Phases I and III, S-3 Rpt, p. 5. 

1 "The account of this attack is from: NTLF 
Rpt Marianas, Phase I, G-2 Rpt, pp. 77-78, 85; 
gist Army Msg File, Msg 1039; 6th Marines SAR 
Saipan, p. 4; Hoffman, Saipan, pp. 86-gi. 

on time. Meanwhile, the marines were able 
to prepare their night positions undis- 
turbed except by artillery and mortar fire. 
About 0330 the Japanese struck-- 
chiefly against the 6th Marines. No less 
than thirty-seven Japanese tanks were in- 
volved, and perhaps a thousand infantry- 
men. They approached the American 
lines through a ravine that cut westward 
through the mountains toward the radio 
station. The tanks came in groups of four 
and five, each with a few riflemen aboard. 
Each group of riflemen carried at least one 
light machine gun. When they came with- 
in range, they were met by a furious bar- 
rage of fire from the marines' artillery, 
machine guns, mortars, bazookas, and rifles. 
Within an hour, a good percentage of the 
tanks had been either destroyed or incapac- 
itated. Although the escorting infantrymen 
kept up the fight until about 0700, their 
efforts were fruitless. By the end of the 
battle the Japanese had lost at least 
twenty-four and possibly more of their 
tanks and an uncounted number of infan- 
trymen. Saito's second counterattack was 
a total failure. 

Change of Plans 

The initial plan for the capture of the 
Marianas had set 18 June as the tentative 
date (W Day) for the landing on Guam, 
which was to constitute Phase II of the 
Forager operation. On the night of 15 
June, after it appeared that the marines 
could hold their narrow beachhead on Sai- 
pan, Admiral Spruance confirmed this 
date, and preparations were set under way 
for an immediate invasion of Guam. But 
before daybreak of the 16th, Spruance re- 
ceived new information that caused him to 
reverse his own decision. 



At 1900 on the evening of 15 June, the 
U.S. submarine Flying Fish sighted a Jap- 
anese task force of battleships, cruisers, 
destroyers, and aircraft carriers making its 
way eastward through San Bernardino 
Strait in the central Philippines. Four 
hours later another submarine, Seahorse, 
reported another enemy task force about 
two hundred miles east of Leytc Gulf 
steaming in a northwesterly direction. 16 It 
was clear that the Japanese Fleet was pre- 
paring to do battle and that the U.S. Fifth 
Fleet would be called upon to take the 
necessary countermcasures. 

The next morning Admiral Spruance, in 
the light of these developments, postponed 
indefinitely the date for the invasion of 
Guam and joined Admiral Turner aboard 
Rocky Mount off the coast of Saipan. To- 
gether, Turner and Spruance decided 
that unloading should continue at Saipan 
through 17 June, that as many transports 
as possible would be retired during the 
night and that only those urgently required 
would be returned to the transport area 
on the morning of the 18th. The old bat- 
tleships, cruisers, and destroyers of the 
Saipan bombardment group would cover 
Saipan from the westward, and Admiral 
Conolly's force would be withdrawn well 
to the eastward out of any presumable 
danger from enemy naval attack. Certain 
cruiser and destroyer units heretofore at- 
tached to Admiral Turner's Joint Expedi- 
tionary Force were to be detached and 
directed to join Admiral Mitscher, who 
would carry the brunt of the attack against 
the approaching enemy fleet. Patrol planes 
based in the Marshalls were to be dis- 

10 Comdr Fifth Fleet, Initial Rpt on Opri to 
Capture the Marianas, 13 Jul 44, p. 3. 

patched forthwith and would prepare to 
make night radar searches as far as 600 
miles west of Saipan. Finally, Admiral 
Mitscher was ordered to discontinue all 
support aircraft operations over Saipan 
and restrict his carrier air operations on 17 
June to searches and morning and after- 
noon neutralization strikes on Guam and 
Rota. Thus were begun the preparations 
for the Battle of the Philippine Sea. 

First Landings of the 
2jth Infantry Division 

The imminence of a full-scale naval bat- 
tle also demanded an immediate decision 
regarding the disposition of the troops of 
the 27th Division, which had been assigned 
to corps reserve. The division had sailed 
from Oahu in three separate transport di- 
visions under command of Rear Adm. 
William H. P. Blandy and was scheduled 
to reach Saipan the day after the main 
landings. On 15 June, while still en route 
to the objective, the 106th Regimental 
Combat Team (RCT) was detached from 
the division and ordered to join Admiral 
Conolly's Southern Attack Force as the 
reserve force for the Guam invasion, which 
at that time was still scheduled to take 
place on 18 June. Shortly before noon of 
the 1 6th, when the ships carrying the other 
two regiments were still about thirty miles 
from Saipan, General Ralph Smith, aboard 
the transport Fremont, was notified by ra- 
dio that the division, less the 106th RCT, 
was to land as soon as practicable over the 
beaches held by the 4th Marine Division. 
The general himself was ordered to report 
to Cambria, flagship of Admiral Hill and 
headquarters of Brig. Gen. Graves B. Ers- 



kine, USMC, chief of staff to Holland 
Smith. 17 

Aboard Cambria, General Ralph Smith 
received his orders to land his division 
artillery as soon as possible to support the 
4th Marine Division. The 165th Regiment 
was to land immediately and move to the 
right flank of the 4th Marine Division, to 
which it would be attached. The 105th 
Regiment would follow. The 106th was to 
remain afloat as reserve for the Southern 
Landing Force for the Guam operation, 
which by now had been postponed indef- 
initely. As soon as the 105th Regiment and 
other elements of the division were ashore 
they were to unite with the 165th and 
relieve the 4th Marine Division on the 
right zone > which included Aslito airfield. 

General Ralph Smith returned to his 
own flagship about 1930, where the assis- 
tant division commander, Brig. Gen. Og- 
den J. Ross, and the 165th Regiment 
commander, Col. Gerard W. Kelley, were 
anxiously awaiting him. Kelley had already 
instructed his executive officer, Lt. Col. 
Joseph T. Hart, to land the regimental 
combat team over Blue Beach 1 immediate- 
ly south of the Gharan Kanoa pier. Ross 
and Kelley were then ordered to go ashore, 
establish contact with the 4th Marine Di- 
vision, and to make whatever arrangements 
were practicable during the night. 

The two officers, accompanied by a 
small advance group, left Fremont about 
2100. The coxswain of their small boat lost 

17 The account of the landing of the 165th Reg- 
iment is derived from: Ralph C. Smith to CG 
USAF Central Pacific Area, Preliminary Rpt On 
Opns of 27th Div at Saipan, 15-24 Jun 44 (11 
Jul 44), Annex I, Notes on Opns of 27th Div at 
Saipan (hereafter cited as Ralph Smith, Notes, 
Saipan), pp. 1-2; Brig Gen Ogden J. Ross, Sum- 
mary of Opns by 27th Div, 16 Jun- 10 Jul ( 13 Jul 
44), pp. 1-2; 165th RCT Rpt of Saipan Action, 
14 Jul 44, pp. 1-2. 

his way, and, after much fumbling in the 
dark and many futile inquiries among 
other landing craft in the area, the party 
finally located a guide boat to steer them 
through the channel to Blue Beach 2, where 
they waded ashore about Q130. |(Mafi //.)| 

In spite of the darkness and confusion 
on the beach, they succeeded in locating 
the command post of the 23d Marines 
about 300 yards south of the point where 
they had landed. General Ross raised 4th 
Marine Division headquarters by telephone 
and was informed that the 165th Regiment 
was expected to move to the right flank of 
the line and jump off at 0730. By this time 
it was 0330 and the Army troops were 
scattered along the beach over a three-mile 
area. General Ross and Colonel Kelley im- 
mediately set forth to locate the command 
post of the 4th Marine Division. There, 
Kelley was ordered by the division chief of 
staff to pass through the lines held by the 
3d Battalion, 24th Marines, and relieve on 
his left elements of the 25th Marines. 
Jump-off hour for the attack toward Aslito 
field was confirmed as being 0730. 

Meanwhile, Colonel Kelley had estab- 
lished telephone contact with his executive 
officer, who reported that the 1st and 
2d Battalions of the 165th Infantry had 
landed. 18 After getting his orders, Kelley 
joined the two battalions and moved them 
south along the road running down the 
beach from Charan Kanoa. Just before 
dawn they took positions along the railroad 
embankment paralleling and east of the 
coastal highway and about 1,000 yards be- 
hind the line of departure. As the first 

1s Thc 3d Battalion, 165th RCT, remained 
afloat during the night. Part of the landing team 
stayed aboard ship because of the scarcity of land- 
ing craft; the remainder spent the night aboard 
landing craft, unable to locate the Charan Kanoa 
channel. 3d Bn 165th RCT Jnl, 16 Jun 44. 



glimpses of light appeared in the eastern 
sky before them, they prepared to jump 
off in support of the 4th Marine Division. 
During these same early morning hours, 
three of the 27 th Division's four artillery 
battalions were also moving toward shore. 
The 105th Field Artillery Battalion landed 
at Blue Bleach 1 at 0515 and by 1055 was 
in position and ready to fire in support of 
the 165th Regiment. The other two field 
artillery battalions (the ro6th and the 
249th) came ashore somewhat later but 
were registered and ready to fire by about 
the same time. The fourth battalion, the 
104th Field Artillery Battalion, remained 
afloat and detached from division ar- 
tillery. 19 

D Plus a: iy June 

165th Infantry 

The immediate objective assigned to the 
165th Infantry, which was attached to the 
4th Marine Division, was Aslito airfield 
and as much of the surrounding area as 
could be secured in a day's fighting. Before 
that could be accomplished, the regiment 
would have to take the small village that 
lay on the boundary line between its two 
battalions, pass through a series of densely 
planted cane fields, and seize the ridge that 
ran in a southwesterly direction along 
most of the regimental front and that com- 
manded the western approaches to the air- 
field. The ridge at its highest points was 
about 180 feet. The distance between the 
line of departure and the westernmost 
point of the airfield along the regiment's 
line of advance was roughly 1,500 yards. 

Colonel Kclley placed his 1st Battalion 
on the right, his 2d on the left. Maj. James 
H. Mahoney, commanding the 1st Battal- 

lu 27th Inf Div Arty Rpt Foragkr Opn, p. 7. 2,1 

ion, disposed B Company on the left, and 
A Company on the right just inshore of the 
southern coast of the island. Lt. Col. John 
F. McDonough put his E Company on the 
right and G Company on the left, tying in 
with the 25th Marines. 20 

The 1st Battalion crossed the line of de- 
parture at 0735, the 2d about fifteen 
minutes later. 21 Company A, on the right, 
immediately ran into a fire fight. Three 
Japanese pillboxes located just inland from 
the beach opened fire on the advancing 
troops and were not eliminated until an 
amphibian tank had been called in to assist 
and engineers were brought up to place 
shaped charges and scorch out the enemy 
inside with flame throwers. 

Along the rest of the regimental line the 
troops ran into no difficulty until they ap- 
proached the small settlement that lay on 
the boundary line between the two bat- 
talions. As B Company tried to skirt south 
of the village, it came under simultaneous 
fire from the direction of the village itself 
and from the ridge to the eastward. 1st Lt. 
Jose Gil, B Company's commander, called 
for an air strike at 0955, but five minutes 
later canceled the request in favor of ar- 
tillery fire from the 14th Marines. 22 

For the next two hours the whole line 
was more or less immobilized. It had be- 
come apparent that the ridge line in front 
was strongly held by the enemy. The ridge 
istelf was covered by sparse undergrowth 
and the approaches to it were all across 
open cane fields. The cane offered some 

20 Unless otherwise noted, this account of the 
action of the 165th Infantry is derived from: 27th 
Inf Div G-3 Jnl; 165th RCT Rpt of Action Sai- 
pan, pp. 2-3; 165th RCT Jnl; 1st Bn 165th RCT 
Jnl; 2d Bn 165th RCT jnl; Edmund G. Love, 
The Battle for Saipan, pp. 43-83, MS in OCMH. 

21 165th RCT Jnl, 17 Jun 44, Msg 13. 
1st Bn 165th RCT Jnl, 17 Jun 44, Msgs 23, 



Narrow-Gauge Railroad Near Charan Kanoa, Infantrymen of the ist and 2d 

Battalions wait for the jump-off signal, ij June. 

cover from enemy observation as long as 
the terrain was level, but entrenched as 
they were on the hill above these fields, the 
Japanese could follow every movement 
made by the Americans approaching below 
them. 23 

By noon Colonel Kelley had more troops 
available. The 3d Battalion, part of which 
had remained aboard its transport while 
the other part spent the night offshore in 
small boats, was finally landed and assem- 
bled during the morning. 24 Company I was 
ordered to report to the ist Battalion com- 
mander to act as reserve in place of C 
Company, which was now to be commit- 
ted to the support of Company B. 

At 1 150 the ist Battalion moved off 
again in the attack with A Company on 
the right, B on the left, and C to the rear 
of B. At 1230 the 2d Battalion jumped off 
following a fifteen-minute artillery prepar- 
ation. 2 '" 1 Immediately, the ist Battalion 
came under a concentration of mortar and 
machine gun fire from the high hill that 
marked the southern extremity of the ridge 
line. For the next hour and fifteen minutes 
this position was pounded by the field 
pieces of the 105th and 249th Field 
Artillery Battalions as well as by naval gun- 
fire. At 1 4 1 4 the attack was resumed. 28 

23 Love, Battle for Saipan, pp. 52-^3. 

24 3d Bn 165th RCT Jnl, t7 Jun 44. 

25 ist Bn 165th RCT Jnl, 17 Jun 44, Msg 32; 
ad Bn 165th RCT Jnl, 16 Jun 44, Msg 39. 

26 ist Bn 165th RCT Jnl, 17 Jun 44, Msgs 39, 
41, 46, 5a, 53, 54, 55. 



Soldiers Watch Destruction of a Pillbox, the last of three that had 
slowed their advance toward Aslito airfield on iy June. 

By 1535 Company A had gained the 
crest after losing three men killed and four 
wounded. 27 About an hour later it was 
joined by two platoons of B Company, but 
the third platoon got involved in a fire fight 
in the cane fields below and failed to reach 
the summit during the rest of the day. 28 

Meanwhile, a gap had developed be- 
tween Companies B and E, and the 1st 
Battalion commander ordered Capt. Paul 
Ryan to pull his C Company around to the 
left of B. Ryan was ordered to make a re- 
connaissance to determine whether he 
could move to the right behind A Com- 
pany and up the ridge by the same route 

21 Ibid., Msg 66. 
28 Ibid., Msg 73 ; 

it had taken. Once on the ridge, it was 
supposed that he could move his company 
directly to the left and take position on the 
left of Company B. Ryan made the crest 
with about half of his second platoon, but 
the rest of his company failed to reach the 
objective. 29 

While Company A and most of Company 
B on top of the hill were digging in and 
Company C was attempting to reinforce 
them by various routes, the Japanese again 
struck. Starting about 1725, the enemy 
managed to work his way between B 
Company and the 2d Battalion and com- 
menced to pound the hill with mortars and 
dual-purpose guns from the southern tip 

Love, Battle for Saipan, pp. 

2n Love, Battle for Saipan, pp. 72-74. 



Reinforcements Moving Inland, Men of the 3d Battalion, 165th Infantry, 
landed on IJ June and proceeded directly to assigned areas. 

of the airfield. 30 After about half an hour 
of this, both Lieutenant Gil and Capt. 
Laurence J. O'Brien, commander of Com- 
pany A, decided to move off the hill, 81 

Captain O'Brien moved over to his ex- 
treme left and ordered his platoons to 
withdraw by leapfrogging. The 3d Platoon 
was to pull back behind the 2d while the 
latter covered, and then the 2d was to pull 
back below the ridge while the 1st covered. 
The 1st Platoon eventually withdrew down 
the hill while O'Brien himself covered its 
movement. The company commander was 
the last man down over the cliff. 

Meanwhile, Captain Ryan, commanding 
Company C, decided to move off to the 
left to reinforce B Company and hold at 
least part of the hill if possible. His attitude 
was reflected by one of his men, Pfc. Cleve 

E. Senor: "I fought all day for this ridge," 
Scnor is reported to have said, "and by 
God I'll help hold it." Both Senor and 
Captain Ryan were killed in the attempt, 
and the C Company platoon joined Com- 
pany A in its withdrawal to the beach. 32 
Captain O'Brien led most of the with- 
drawing battalion back along the southern 
beach for a distance of about 1,400 yards, 
then cut inland where he met guides from 
battalion headquarters. Shortly after 2000 
he reached the command post with ele- 
ments of all three companies and dug in 
for the night practically at the line of 
departure from which the companies had 
attacked in the morning. Except for scat- 
tered elements that remained dug in along 
the approaches to the ridge, progress in 
the 1 st Battalion's zone of action had been 

30 1 st Bn 165th RCT Jnl, 17 Jun 44, Msg 79. 
31 /fc^,, Msgs 80, 81. 

'' 2 Love, Battle for Saipan, pp. 75-77. 



nil. 33 Casualties for the day's fighting in 
the battalion were reported as 9 killed and 
21 wounded. 34 

The 2d Battalion had been more success- 
ful. After the 1230 jump-off, E Company, 
on the battalion right, was immediately hit 
by an enemy artillery barrage that killed 
three men and wounded four others. Ex- 
cept for the 1 st Platoon, the whole com- 
pany retired to the extreme west edge 
of the village that lay on the battalion 
boundary line and for the next hour 
reorganized its scattered elements and evac- 
uated its wounded. The 1st Platoon, how- 
ever, instead of withdrawing when the 
artillery barrage hit, rushed forward in an 
effort to take concealment in the heavy 
cane at the foot of the ridge line. From 
there it began to move on to the ridge it- 
self, but after the leading squad was cut 
off by Japanese fire, the rest of the platoon 

Capt. Bernard E. Ryan, the company 
commander, had been with the forward 
elements of the 1st Platoon when his com- 
pany was hit and was already in the cane 
field making a reconnaissance forward. 35 
With two of his men, he made his way 
through the cane and up to the top of the 
ridge. For thirty minutes they waited in 
vain for the rest of the platoon to come 
up, and when it finally appeared that they 
were isolated, Ryan decided to conduct a 
reconnaissance. For three hours this officer 
and his two men wandered around the 
hilltop observing the enemy from a distance 

33 Ibid., p. 79. 

:il 27th Inf Div G-i Periodic Rpt 1. 

35 Ryan was the brother of Capt. Paul Ryan, C 
Company commander, who was killed later in the 
afternoon while trying to hold a portion of this 
same ridge. Because of Paul Ryan's death, men 
of the 27th unofficially named the position Ryan's 

sometimes of only thirty yards. He ordered 
one of his men, S. Sgt. Laurence I. Kemp, 
to carry the information gained back to the 
company executive officer. Kemp, equally 
fearful of friendly and enemy fire along the 
return route, solved his dilemma by tying 
a white hankerchief to the barrel of his 
rifle, executing a right shoulder arms, and 
marching safely down the hill in full view 
of both the enemy and his own troops. 

Upon receiving Kemp's information 
the battalion commander immediately re- 
quested reinforcements. Colonel Kelley re- 
leased F Company, which was then moved 
into the line to the left of E. Both com- 
panies jumped off at 161 o behind a screen 
of heavy mortar, small arms, and auto- 
matic weapons fire. 3fi Within thirty min- 
utes they reached the ridge line about two 
hundred yards west of Aslito field and 
began to dig in. 37 

On the extreme left of the battalion 
front, Capt. Paul J. Chasmar's G Company 
met with little difficulty. By 141 6, less than 
two hours after the jump-off, the com- 
pany had reached the ridge line and 
commenced to dig in. 38 Chasmar sent two 
patrols onto the airfield. They investigated 
the installations along the west side of the 
field and up to the south edge of the 
stretch without running into opposition. 
About 1530 temporary contact was estab- 
lished on the left with the 25th Marines, 
which had by this time penetrated into 
the building area north of the airfield 
proper. yfl 

Thus, by the end of 17 June the 2d 
Battalion had succeeded in pushing about 

86 2d Bn 165th RCT Jnl, 17 Jun 44, Msg 51. 

ST Ibid., Msg 53 ; Love, Battle for Saipan, pp. 

S8 ad Bn 1 65th RCT Jnl, 17 Jun 44, Msg 46. 

3 ' J Love, Battle for Saipan, p. 60; Hoffman, 
Saipan, p. 95. 



1,300 yards from the line of departure, was 
firmly dug in just two hundred yards short 
of Aslito airfield, and was in a good posi- 
tion to attack the field the following 
morning. In the day's fighting the bat- 
talion had lost six killed and thirty-six 
wounded. 40 It failed to attack the airfield 
on the 17th only because of regimental 
orders to the contrary. Colonel Kelley de- 
cided that in view of the difficulty encoun- 
tered by his 1st Battalion on the right 
flank, it would be unwise for the 2d Bat- 
talion to push forward any farther. From 
its positions on top of the ridge line 
commanding Aslito, the 2d Battalion "had 
an excellent field of fire against any pos- 
sible counterattack," so the regimental 
commander ordered it to hold there for 
the night and to resume the attack against 
the airfield the next day. 41 

4th Marine Division 

To the left of Colonel Kelley's regiment 
the 25th Marines jumped off at approxi- 
mately the same time in columns of battal- 
ions. Against light resistance the regiment 
pushed rapidly ahead to its 0-2 line. 
Because of the marines' more rapid prog- 
ress, a gap developed between them and 
the 2d Battalion, 165th Regiment, that was 
filled by two companies of marines. By 
midafternoon the companies had searched 
the building area north of the airfield 
proper and sent patrols onto the field itself. 
When Colonel Kelley's determination not 
to attack the airfield until the 1 8th became 
known to the 25th Marines, its 3d Battalion 
was shifted to the north side of the airfield, 
facing south, and as it dug in for the night 

there was no contact between the marines 
and the Army unit. 42 

In the center of the 4th Marine Divi- 
sion's line, progress was more difficult. The 
24th Marines jumped off on time about 
0730. In spite of continuous fire from 
antiaircraft guns located east of the air- 
field, the right flank battalion reached the 
foot of the ridge line quickly and by noon 
commenced the ascent. By 1630 the bat- 
talion commander reported that his men 
were digging in on the O-2 line. In the 
center and to the left enemy resistance was 
even stronger, and after reaching the ap- 
proaches to the ridge by late afternoon, the 
marines withdrew a full 600 yards before 
digging in for the night. 43 

To the 23d Marines on the division's left 
flank fell the hardest fighting in the 4th 
Marine Division zone for the 1 7th. On the 
right, the 2d Battalion made fairly rapid 
progress against light opposition, but on 
the left, the 1st Battalion was not so 
fortunate. Having once cleared Fina Susu 
ridge, the marines started to advance 
across the open ground to the eastward but 
were quickly pinned down by heavy mortar 
and enfilade machine gun fire from their 
left front. After retiring to the ridge line 
to reorganize, the battalion pushed off 
again at 1500 after a ten-minute artillery 
fire. Again the attack was stopped. Mean- 
while, the 2d Battalion on the right had 
been pushing steadily forward .and contact 
was lost between the two battalions. Even 
more serious was the 600-yard gap on the 
left between the 23d Marines and the right 
flank of the 2d Marine Division. From this 

40 27th Inf Div G-i Periodic Rpt 1. 

41 165th RGT Rpt of Action Saipan, p. 3. 

42 4th Marine Div Opns Rpt Saipan, Annex J, 
25th RGT Rpt, p. 4; Hoffman, Saipan, p. 95. 

4:1 4th Marine Div Opns Rpt Saipan, Annex I, 
24th RCT Rpt, pp. 18-19; Hoffman, Saipan, pp. 



MAP 3 

area came most of the enemy fire, and the gered the flanks of both. 

failure of the two Marine divisions to close As night approached it became apparent 

this gap early in the day seriously endan- that, with the advance of the 2d Battalion 



* m*""*~i i*&i 

105TH Infantrymen Wading in From the Reef on 17 June. 

and the delay of the 1st, the right flank 
was extended and the left retarded so that 
it was impossible to close the gap with the 
units then on line. Consequently, the 3d 
Battalion, 23d Marines, was ordered to tie 
in the flanks of the two. Later, the 3d 
Battalion, 24th Marines, was attached to 
the 23d Regiment and under cover of dark- 
ness was moved into position to relieve the 
3d Battalion, 23d Marines, tie in, and de- 
fend the gap between the two leading 
battalions. But between the two Marine 
divisions as they dug in for the night, the 
wide gap in the area around Lake Susupc 
still remained unclosed. 44 

2d Marine Division 

In the zone of the 2d Marine Division, 
the day's plan called for an attack by the 
2d and 6th Regimental Combat Teams to 

the northeast, while the 8th Marines, on 
the division right, was t o drive due east 
toward the O-i line. 45 \{Map g)\ The 
jump-off hour was originally scheduled for 
0730 but was subsequently changed by 
General Holland Smith's headquarters to 
0930. Word of the change, however, failed 
to reach division headquarters in time, so 
the troops crossed the line of departure ac- 
cording to the original schedule, following 
a 90-minute intensive preparation by aerial 
bombardment, naval gunfire, and artillery 

On the extreme right, the marines of the 
2d Division met with the same problems 
that were besetting the left flank of the 
4th Division, and more besides. The 1st 
Battalion, 29th Marines, attached to the 
8th Marines, had first to slosh its way 

44 4th Marine Div Opus Rpt Saipan, Annex H, 
23d RCT Rpt, pp. 36-37; Hoffman, Saipan, p. 96. 

4 "The following account of action in the 2d 
Marine Division zone is from: 2d Marine Div 
SAR, Phase I, Foragkr, Sec, VI, p. 4; Hoffman, 
Saipan! pp. 92-93. 



through the sniper-infested swamp that 
ran about 1,000 yards north of Lake 
Susupe. 4fi Directly east of the swamp was 
a coconut grove from which periodically 
came enemy mortar fire, described in the 
division action report as "bothersome." 
Northeast of the coconut grove was a 
high hill on which the Japanese were 
entrenched in caves, and beyond this on a 
sharp nose was a series of heavily manned 

Throughout the day the ist Battalion, 
29th Marines, was unable to seize the 
coconut grove and in fighting for it the 
battalion commander, Lt. Col. Guy E. 
Tannyhill, was wounded and had to be 
evacuated. By late afternoon the battalion, 
with the help of four tanks of the 2d Ma- 
rine Tank Battalion, succeeded in taking 
the hill to the north of the grove where 
it dug in for the night. Meanwhile, the 
other two assault battalions of the 8th Ma- 
rines had reached their objective line with 
little difficulty and were tied in for the 
night with the 6th Marines on their left. 

The 6th Marines had jumped off on 
schedule at 0730 and soon after ogoo had 
reached its objective line, encountering 
little resistance on the way. Further prog- 
ress was held up because of the danger of 
overextending its lines as a result of the 
relatively slow progress of the 8th Marines 
on the right. 

The 2d Marines, on the division left, 
regulating its advance by that of the reg- 
iment to its right, moved forward at 0945 
in a column of battalions. By 1020 the 
leading battalion had advanced four hun- 
dred yards against light resistance. By 1800 

the regiment had reached its objective line, 
which was coincident with the Force 
Beachhead Line 47 in its zone and lay only 
a thousand yards from the southern out- 
skirts of the town of Garapan. 

Landing Reinforcements 

At 0605 on the 17th, Col. Leonard A. 
Bishop received orders to land his 105th 
Regimental Combat Team as soon as boats 
were available. 48 By 0845 the ist Bat- 
talion was loaded and headed for the 
beach; the other two followed during the 
morning. 48 However, because of low tide 
and the heavy congestion in and around 
the Charan Kanoa channel, the troops had 
to be landed piecemeal. Not until late 
afternoon were all of the infantrymen 
ashore. That evening the 2d Battalion was 
attached to the 4th Marine Division as re- 
serve, and the ist Battalion was attached 
to the 165th Infantry and moved to an 
assembly area just west of Aslito field. 
Also, the 27th Division Reconnaissance 
Troop landed and commenced to establish 
an observation post area running from 
Agingan Point about 1,500 yards along the 
southern shore. The rest of the 105th Reg- 
iment remained in bivouac in the area of 
Yellow Beach 3 during the night. 50 

The slowness with which the 1 05th Reg- 
iment was landed brought one later em- 
barrassment to that unit. In view of the 
bottleneck at the Charan Kanoa channel, 
orders were issued shortly after noon to 
stop unloading equipment through the 
channel until the congestion had been 

40 The swampy ground around Lake Susupe was 
much more extensive than it had appeared on the 
map used by U.S. planners. See map of Saipan 
Island reproduced from captured Japanese maps 
by ACofS G-2 NTLF, 26 Jun 44. 

47 The Force Beachhead Line is the line that 
fixes the inshore limits of a beachhead. 
18 105th RGT Jnl, 17 Jun 44, Msg 4. 

49 Ibid., passim. 

50 Ibid., Msgs 60, 61, 67; 105th RCT Opn Rpt 
Forager, pp. 4-5. 



cleared up. 51 This caught most of the 
regiment's organizational equipment still 
aboard the transport Cavalier. That night 
Cavalier, along with most of the other 
transports, retired eastward after an air raid 
warning. Meanwhile, the Japanese fleet 
was reported to be moving toward Saipan, 
In the light of these circumstances, Cava- 
lier was ordered to stay out of the danger 
zone and did not return until 25 June to 
continue unloading.* 2 As General Ralph 
Smith later testified : 

The 105th Infantry was thus placed under 
great handicap in operating as a regimental 
unit. It had very little communication equip- 
ment or personnel ashore, cither radio or tel- 
ephone. It had almost no staff facilities or 
blackout shelter such as regimental head- 
quarters is compelled to use if orders arrive 
after dark, 53 

North of the 27th Division's beaches 
other important elements were coming 
ashore on the 17 th. General Holland Smith 
left Rocky Mount in midafternoon and at 
1530 set up the Northern Troops and 
Landing Force command post at Charan 
Kanoa. General Harper, corps artillery 
commander, moved his command post to a 
point about 200 yards inland from Yellow 
Beach 2, and advance parties of the 532d 
Field Artillery Battalion got ashore. 54 

Night of 1 J- 18 June 

Compared to the first night on Saipan, 
that of the 17th was quiet for the Ameri- 
can troops in their foxholes. Only in the 
zone of the 2d Marine Division did the 

" 27th Inf Div G-4 Jnl, 17 Jun 44. 

S2 USS Cavalier, Action Rpt Saipan-Tinian, 28 
Aug 44, p. 3. 

5S Ralph Smith, Notes, Saipan, p. 3. 

H1 NTLF Rpt Marianas, Phase I, p. 12; XXIV 
Corps Arty Final Rpt on Forager Opn, Phases I 
and III, S-3 Rpt, pp. 5-6. 

Japanese exert themselves. Around mid- 
night, they attempted to breach the Ma- 
rine lines near the boundary between the 
6th and 8th Regiments. About fifteen or 
twenty Japanese overran two machine 
guns, but the attack was shortly stopped. 
For a brief time the enemy penetration 
destroyed contact between the two regi- 
ments, but the gap was quickly filled and 
the lines were restored. 55 

A more serious enemy threat occurred 
on the morning of the 18th in the form of 
an attempted counteramphibious landing. 
A month before the American landings, 
31st Army had established a force consist- 
ing of the 1st Battalion, 18th Infantry, to 
be held in readiness for amphibious attacks 
on either Saipan or Tinian in the event the 
Americans were able to establish a beach- 
head." 6 About 0430 on the 18th this group 
sorticd from Tanapag Harbor in thirty-five 
small boats to put the plan into effect. The 
Japanese failed. LCI gunboats intercepted 
the boats and, with the help of Marine 
artillery, destroyed most of the landing 
party and turned back the rest. 57 

This uninterrupted scries of reverses sus- 
tained by the Japanese on Saipan merely 
reinforced their determination to hold the 
island at all costs. On the t 7th the chief 
of the Army General Staff in Tokyo at- 
tempted to bolster the spirits of the defen- 
ders in a message to 31st Army headquar- 
ters: "Because the fate of the Japanese 
Empire depends on the result of your op- 
eration, inspire the spirit of the officers and 
men and to the very end continue to 
destroy the enemy gallantly and persistent- 

* r ' Hoffman, Saipan, pp. 99-100, 

s,i JICPOA Trans 8905, SONAE Operation Or- 
der 44 ( 14 May 44) 31st Army Order. 

57 NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, p. 14. 



ly; thus alleviate the anxiety of our 

To which the Chief of Staff, 31st Army, 
responded: "Have received your honorable 
Imperial words and we are grateful for 
boundless magnanimity of the Imperial 
favor. By becoming the bulwark of the Pa- 
cific with 1 0,000 deaths we hope to requite 
the Imperial favor," 58 

D Plus 3: 18 June 
27th Division 

General Holland Smith's orders for 18 
June called for all three divisions under his 
command to seize the O-3 line within their 
respective zones of action. For the 4th Ma- 
rine Division and the 27th Division this 
meant that the end of the day should sec 
them resting on the eastern coast of Saipan 
from a point opposite Mount Nafutan up 
the shore line about 5,000 yards in a north- 
erly direction to a point about one third 
up Magicienne Bay. From there the ob- 
jective line for the 4th Marine Division 
bent back in a northwesterly direction to 
correspond with the advance of the 2d Ma- 
rine Division, which was not intended to 
cover so much territory. The boundary be- 
tween the 4th Marine Division and the 
27th Infantry Division ran eastward to 
Magicienne Bay, skirting Aslito field to the 
north. Army troops were to capture the 
field itself 59 |(S7"W//.) I 

For action on the 18th, the 27th Division 
had under its command only the 165th 
Regiment and the 1st and 3d Battalions of 
the 105th, The 2d Battalion, 105th Regi- 
ment, remained in corps reserve in an area 
to the rear of the 4th Marine Division, and 

DS gist Army Msg File, Msgs 115 and 1046. 
59 NTLF Opn Order 6-44, 1 7 Juii 44. 

the 106th Infantry was still at sea. In spite 
of the fact that as early as 0758 the Marine 
division had notified General Ralph Smith 
that control of the 165th Regimental Com- 
bat Team was passing to Col. Kclley,* 50 the 
regimental commander remained uncertain 
as to his own exact status. He later re- 
ported : 

I was unable to determine (by telephone 
conversation with Hq 4th Marine Div) 
whether I was still attached to the 4th Ma- 
rine Division or had passed to the command 
of CG 27th Div. . . . Shortly after this, Ma- 
jor General Ralph Smith visited my CP and 
advised me that I should receive notice of 
my release from the Marines and reversion to 
the 27th Division. I did receive notice from 
the 27th Division but never received such 
orders from 4th Marine Division Head- 
quarters. 61 

This confusion, however, though indica- 
tive of poor liaison, was to have no 
significant effect on the action of the units 

Jump-off hour for the two Marine di- 
visions was to be 1000; for the Army 
division it was 1200. 62 The immediate 
concern of Colonel Kellcy, however, was to 
recapture the ridge southwest of Aslito 
that his 1st Battalion had given up the 
previous day. Accordingly, at 0605, he or- 
dered Maj, Dennis D. Claire to move the 
3d Battalion into the line on the right in 
order to launch a co-ordinated attack with 
the 1 st Battalion at 07 30. 03 The 165th In- 
fantry jumped off on schedule after a 
half-hour naval and artillery preparation 
along the whole front. The ist and 3d Bat- 
talions with four tanks preceding them 
stormed up the ridges while the 2d Bat- 

so a 7th Inf Div G-3 Jul, 18 Jim 44, Msg 16. 

liL 165th RCT Rpt Saipan, p. 3. 

52 NTLF Opn Order 6-44. 

153 3d Bn 165th RCT Jnl, 18 Jun 44, Msg 3. 



talion on the edge of the airfield held its 
lines until the other units on its right came 
abreast, A few minutes after iooo the 
ridge that had caused so much trouble the 
preceding day was secured against very 
light opposition and with negligible casual- 
ties to the assaulting units/' 4 

Meanwhile, at 0800, Colonel Kelley au- 
thorized his 2d Battalion to cross Aslito 
airfield. 63 Beginning about 0900, Captain 
Chasmar, commanding G Company, or- 
dered his men across the airfield along the 
north side. Capt. Francis P. Leonard, in 
command of F Company, followed suit 
shortly after, although he kept his com- 
pany echeloned to the right rear in order to 
keep physical contact with E Company, 
which in turn was in contact with the 1st 
Battalion. Chasmar reported that he had 
crossed the airfield at 1000. Sixteen min- 
utes later, Aslito was announced as se- 
cured,* 56 That afternoon when General 
Ralph Smith arrived at the regimental 
command post the airfield was officially re- 
named Conroy Field in honor of Col. 
Gardiner J. Conroy, former regimental 
commander of the 165th, who had been 
killed at Makin. 67 Later, it was renamed 
Iselcy (sic) Field in honor of a naval avi- 
ator, Comdr. Robert H. Isely, who had 
been shot down over Saipan/' 8 

Up until 1000 the troops that had over- 
run the airstrip had met no opposition. 
Only one Japanese was discovered on the 
whole installation, and he was found hiding 
between the double doors of the control 

,i4 165th RGT Jnl, 18 Jun 44, Msgs 45, 54, 
«■> Ibid., Msg 16. 

e6 Ibid., Msg 56; Love, Battle for Saipan, pp. 

67 165th RGT Jul, 18 Jun 44, Msg 76. See 
Growl and Love, Gilberts and Marshalts, pp. 96-97. 

68 Hoffman, Saipan, p. 10411. 

tower. All of the Aslito garrison still alive 
had retired to Nafutan peninsula. 60 

Upon reaching the eastern end of the 
airstrip, Captain Chasmar stopped to build 
up his line because he had been having 
considerable trouble during the morning 
trying to cover his frontage. He had tried 
unsuccessfully to make contact with the 
marines on the left who were now veering 
off to the northeast and in his move across 
the airport had temporarily lost contact on 
the right with F Company. At the same 
time, F Company was itself developing 
large gaps between platoons. By 1100 the 
whole 2d Battalion advance was stopped 
while the battalion commander waited for 
his companies to close up. For the next 
two hours the forward line remained sta- 
tionary along the eastern boundary of the 
airfield. Unfortunately, the terrain in 
which G and F Companies had taken up 
positions was overlooked by the high 
ground of Nafutan ridge, and the men 
had hardly begun to dig in when they 
came under fire from dual-purpose guns 
located in that sector. The fire lasted for 
about two hours until friendly artillery was 
brought to bear on the Japanese positions, 
which were temporarily silenced. 70 

With the airfield secure in the hands of 
the ad Battalion, 165th Regiment, and the 
ridge west and southwest of it occupied 
by the 1st and 3d Battalions, General 
Ralph Smith rearranged his units to launch 
the main attack at noon as ordered. Into 
his right flank he ordered the 1st and 
3d Battalions, 105th Regimental Combat 
Team, which had landed the day before 
and so far had seen no action on Saipan. 
On the extreme right the 3d Battalion, 

09 Love, Battle for Saipan, p. 87. 
70 ad Bn 165th RCT Jnl, 18 Jun 44; Love, 
Battle for Saipan, pp. 87-88. 



Aslito Field Becomes Conroy Field. General Ralph Smith (left) congratu- 
lating Col. Gerard W. Kelley at ceremonies renaming the airfield following its cap- 
ture. Brig. Gen, Ogden Ross, at right, looks on. 

105th, completed the relief of the 3d Bat- 
talion, 165th, at 1245, three quarters of an 
hour late. 71 The 3d Battalion, 165th, then 
went into reserve. About the same time, 
the 1 st Battalion, 105th, relieved the 1st 
Battalion, 165th. The latter was then 
shifted to the left flank of the division line 
to close the gap between the 4th Marine 
Division and the 2d Battalion, 165th, 
which was occupying the airfield. 73 From 
right to left, then, the new division line 
consisted of Companies L, I, C, and A, 

71 3d Bn 165th RCT Jnl, 18 Jun 44, Msgs 34, 


73 ist Bn 165th RCT Jnl, 18 Jun 44, Msgs 41, 

43= 44. 45. 57, 

105th Infantry, and Companies F, G, B, 
and C, 165th Infantry, with the remaining 
infantry companies in reserve in their re- 
spective battalion zones. 

As the afternoon wore on it developed 
that, as on the previous day, progress on 
the extreme right of the division front was 
the slowest, and again the chief obstacle 
was terrain. In the area inland from the 
southern coast the ground was a series of 
jagged coral pinnacles that jutted up from 
the water's edge to a height of about 90 
feet. Between these peaks were a heavy un- 
dergrowth of vines, densely planted small 
trees, and high grass. Against these odds, 
but luckily not against the added encum- 



brance of Japanese opposition, Company 
L on the extreme right advanced a mere 
200 yards from the line of departure by 
nightfall, and I Company's progress was 
only a little better. 73 The situation on the 
left of the 105th Regimental line was some- 
what more promising. In spite of artillery 
fire from Nafutan Point, Lt. Col. William 
J. O'Brien, 1st Battalion commander, suc- 
ceeded by 1400 in pushing forward to a 
line running southwest from the southeast 
corner of the airfield. 74 

While the 105th Regiment was having 
more than a little difficulty getting started 
on the division right flank, the 165th on the 
left was faring better. By 1700 the entire 
regimental line had almost reached Ma- 
gicienne Bay, having met only light opposi- 
tion. The original intention had been to 
proceed on to the water's edge, but the 
heavy undergrowth and coral outcroppings 
persuaded the regimental commander to 
pull back to the high ground west of the 
shore line for the night. About 1700 the 
commanding officer of the 3d Battalion, 
25th Marines, who was on the right flank 
of the 4th Marine Division line, reported 
the imminence of a Japanese counterattack 
between the 24th and 25th Marines. In 
view of the necessity of the hitter's pulling 
north to pour in reinforcements against 
this threat, the lines of the 165th were 
shifted left about 600 yards to establish 
contact with the marines for the night. 75 

4th Marine Division 

North of the 27th Division zone, the 4th 
Marine Division attacked toward the east 

coast with three regiments abreast: the 
25th Marines on the right, 24th Marines in 
the center, and 23d Marines on the left. 76 
The right half of the objective line for this 
day's action was to be on the coast of 
Magicienne Bay, and from there it bent 
back to the northwest to meet the more 
slowly progressing 2d Marine Division. 

The 25th Marines jumped off on sched- 
ule at 1000. Opposition was light, and by 
1330 the regiment had reached the beaches 
on Magicienne Bay well in advance of the 
165th Infantry on its right. The occupation 
of these beaches on the cast coast completed 
the initial drive of the division across Sai- 
pan. The island now, at this point at least, 
was cut in two. One battalion of the 25th 
Regiment was left behind to mop up the 
southern extremity of a heavily defended 
cliff line that had been bypassed by the 
24th Marines on its left. 

The latter regiment had had a little dif- 
ficulty organizing its lines before the jump- 
off and consequently was delayed forty-five 
minutes in the attack. Nevertheless, in the 
face of "moderate to heavy" machine gun 
and rifle fire, it had succeeded by 1400 in 
pushing forward to a point only 300 yards 
west of Magicienne Bay. Then, about 16 15, 
two Japanese tanks suddenly appeared in 
the zone of the 2d Battalion, causing con- 
siderable anxiety and about fifteen Ameri- 
can casualties before they were chased 
away by bazookas and artillery. By night- 
fall the elements of the 24th on the right 
had reached the O-3 line, part of which 
rested on the coast, and the unit was well 
tied in with the regiments on its right and 

73 Love, Battle for Saipan, pp. 89-90. 

74 105th RCT Jnl, 18 Jun 44, Msg 35. 

75 165th RCT Jnl, 18 Jun 44, Msgs 93, 99, 100, 

70 This account of 18 June action of the 4th 
Marine Division is derived from: 23d RCT Rpt, 
PP. 37-38; 24th RCT Rpt, p. 19; 25th RCT Rpt, 
pp. 4-5; Hoffman, Saipan, pp. 102-04. 



Examining an Enemy Gun. This Japanese Type 10 120-mm. gun was one of 
several captured when Aslito airfield was overrun. 

As was the case of the 165th Infantry, 
the 23d Marines on the extreme left of the 
4th Division line had to capture its line 
of departure before the scheduled jump-off 
hour. At 0730 the 3d Battalion, 24th Ma- 
rines (attached to the 23d), passed 
through the 1st Battalion, 23d, with orders 
to seize the line of departure before the 
main attack, which was scheduled for 
0900. The battalion never made it — not 
that day at least. Intense mortar and en- 
filade machine gun fire from the left flank 
stopped the men after an advance of about 
200 yards. On the right the 2d Battalion, 
23d, made about the same gain before it 
too was pinned down. At 1300 the attack 

was resumed and after fierce fighting 
against stubborn Japanese resistance the 
troops advanced about 300 yards. By 17 15 
the regiment had established a line some 
400 yards east of Lake Susupe. Progress 
on the left flank of the 4th Marine Division 
had been far less than anticipated. It was 
becoming apparent that the main line of 
Japanese resistance would be in the area 
north and cast of Lake Susupe and not in 
the southern sector of the island. 

2d Marine Division 

The left flank of the corps line remained 
almost stationary during the 18th. As the 



2d Marine Division's commander explained 
it, "At this stage, the frontage occupied by 
the Division was such that its lines could 
not be further lengthened without danger- 
ously thinning and overextending them." 77 
The strong pocket of resistance encoun- 
tered by the 4th Division near the division 
boundary line formed a hostile salient into 
the beachhead and forced both divisions 
to maintain abnormally long lines in the 
sector. The inability of the 4th Division to 
make substantial progress on its left flank 
in turn prevented the 2d Division from 
risking further extension of its own lines. 
Only the 8th Marines saw significant 
fighting on the 18th. The enemy-infested 
coconut grove on the regiment's right that 
had proved so bothersome the previous day 
was assaulted and captured. Here, a large 
number of Japanese dead were found. Be- 
fore the 1 8th the enemy had systematically 
removed its dead before the advance of the 
attacking forces, but by now, with the 
American beachhead firmly established and 
Aslito airfield overrun, Japanese comman- 
ders on Saipan had more urgent matters 
on their minds. Tft 

The Japanese Situation 

By the night of 18 June, the Japanese 
high command in Tokyo as well as its sub- 
ordinates on Saipan were at last compelled 
to confess that the situation was critical. 
The island had been cut in two and the 
southern part, including the main airfield, 
was for all practical purposes in American 
hands. True, remnants of Japanese units, 

including most of the Aslito garrison, were 
still holed up on Nafutan peninsula and 
along the southern shore west of it, but 
they were cut off from the main body of 
troops and incapable of anything more ser- 
ious than harassing attacks against the 
American lines. | [See Map £Y\ 

In the face of unrelenting pressure from 
their attackers, the Japanese on the 18th 
began withdrawing to a defense line ex- 
tending across the island in a southeasterly 
direction from a point just below Garapan 
via the south slopes of Mount Tapotchau 
to Magicienne Bay. To be more exact, the 
new "line of security" drawn up by 31st 
Army headquarters on the night of the 
1 8th was to run from below Garapan east 
to White Cliff, then south to Hill 230 
(meters) and southeast through Hill 286 
(meters) to a point on Magicienne 
Bay about a mile west of the village of 
Laulau. 79 

The line roughly paralleled the O-4, or 
fourth phase line of the American attack- 
ers, and was the first of two last-ditch de- 
fense lines scratched across the island in a 
vain attempt to stabilize the battle during 
the retreat to the north. If the Americans 
could be brought to a standstill, the Japa- 
nese hoped to prolong the battle and 
eventually win out with the aid of rein- 

77 2d Marine Div SAR, Phase I, Forager, Sec. 
VI, p. 6. 

78 Hoffman, Saipan, p. roa. 

78 CINCPAC-CINCPOA Item 9983-85, P- 10, 
31st Army CofS, Msg 1060. The map of Saipan 
prepared before the operation by American troops, 
on which the maps in this volume arc essentially 
based, is by necessity not very accurate — particu- 
larly from the point of view of terrain configura- 
tion and spot elevation data. 

The troops also used a captured Japanese map 
of Ihc island. The heights on this map, which is 
more accurate as to relief features, are in meters. 
Elevations in meters derived from the Japanese 
map are shown on the maps of this volume in 



forcements. To Tokyo, 31st Army head- 
quarters radioed its plans: 

Situation evening of 1 8 June : 

The Army is consolidating its battle lines 
and has decided to prepare for a showdown 
fight. It is concentrating the [43d Division] 
in the area E of Tapotchau, The remaining 
units [two infantry battalions of the 135th 
Inf, about one composite battalion, and one 
naval unit], are concentrating in the area E 
of Garapan. This is the beginning of our 
showdown fight. 80 

In reply, Imperial General Headquarters 
ordered Maj. Gen. Keiji Iketa to hold on 
to the beaches still in his possession, wait 
for reinforcements over those beaches, and 
"hinder the establishment of enemy air- 
fields,'"* 1 Iketa reported that he would 
carry out these orders, that Aslito airfield 
would be neutralized by infiltration pa- 
trols "because our artillery is destroyed," 

and that the Banaderu (Marpi Point) air- 
field would be repaired and defended "to 
the last." "We vow," he concluded, "that 
we will live up to expectations." 82 Stabil- 
ize the battle, keep beaches open for 
reinforcements, recover and preserve the 
use of the Marpi Point airfield, and deny 
to the Americans the use of Aslito — these 
four objectives were now the cornerstones 
of Japanese tactics on Saipan. 

And from the Emperor himself came 
words of solemn warning and ominous 
prescience: "Although the front line offi- 
cers are fighting splendidly, if Saipan is 
lost, air raids on Tokyo will take place 
often, therefore you absolutely must hold 
Saipan." 83 

Five months later American B-29 bomb- 
ers taking off from Saipan for Tokyo would 
confirm the Emperor's worst fears. 

so 31st Army Msg File, Msg 1050. 
81 Ibid., Msg 150. 

Ibid., Msg 1054. 
Ibid., Msg 152. 


Supporting Arms and Operations 

Battle of the Philippine Sea 

While the marines and soldiers of the 
V Amphibious Corps were still pushing 
toward the east coast of Saipan, Admiral 
Spruance received news that was to prove 
even more significant than that of the cap- 
ture of Aslito field. Beginning on 15 June, 
American submarines patrolling the waters 
east of the Philippine Islands sent in a ser- 
ies of reports on enemy ship movements 
that seemed to indicate strongly that the 
Japanese were massing a fleet and were 
sending it to the rescue of the beleaguered 
defenders of Saipan. 

Spruance, like all other high-ranking 
U.S. naval commanders in the Pacific, had 
hoped that an invasion of the Marianas 
would bring the enemy fleet out fighting. 
That hope now seemed likely to be ful- 

The Japanese Navy, like the American, 
had long been imbued with Alfred Thayer 
Mahan's doctrine that the sine qua non 
of victory in naval warfare is the destruc- 
tion of the enemy fleet. In their own na- 
tional history, the Japanese had only to 
look back as far as 1905 for historical 
warrant for this assumption. In that year, 
Admiral Hcigachiro Togo had met and al- 
most annihilated the Russian Fleet at the 
Battle of Tsushima, thus paving the way to 
Japan's victory in the Russo-Japanese War 

and a long-coveted place in the interna- 
tional sun. 

In the early stages of World War II the 
Japanese sought to put this doctrine to 
test, but always fell short of complete 
success. In spite of the tremendous damage 
done to it at Pearl Harbor, the U.S. Fleet 
survived and recovered with remarkable 
rapidity. Important sea battles were won 
by the Japanese in the Solomons, but none 
of them were conclusive. At Midway the 
tables were turned when U.S. carrier 
planes repelled an attempted invasion and 
administered a sound drubbing to the Jap- 
anese naval forces supporting it. 

By late 1943 the high command of the 
Imperial Navy felt that conditions were 
ripe for a decisive fleet engagement. Twice 
in the autumn of that year Admiral 
Mineichi Koga, Commander in Chief, 
Combined Fleet, sallied forth from Truk in 
an effort to engage the U,S, Central Pa- 
cific Fleet. Both times he failed to discover 
his adversary. In the end he retired to Truk 
and allowed most of his carrier air strength 
to be diverted to the Rabaul area, where 
two thirds of it was lost, 1 In the spring of 
1944, as American forces threatened to 

1 Thomas Wilds, "The Admiral Who Lost His 
Fleet," United Statex Naval Institute Proceedings, 
Vol. 77, No. 11, Whole No. 585 (November, 
'95')j PP- n75-0i; Crowl and Love, Gilberts 
and Marshalls, Ch. IV; Morison, Aleutians, Gil- 
berts and Marshalls, Ch. IX. 



press farther into western Pacific waters, 
the Japanese prepared another plan, Oper- 
ation A-GOj in the hope of forcing a major 
fleet engagement. 

On 3 May 1944 Admiral Toyoda, Koga's 
successor as Commander in Chief, Com- 
bined Fleet, issued the general order for 
Operation A-GO. It was assumed that the 
next major thrust of the U.S. Fleet would 
be into waters around the Palaus, in the 
western Carolines, and that there it could 
be met and bested by the Japanese. 
Thought was given to the possibility that 
the Americans might move first against the 
Marianas rather than Palaus, but the con- 
sensus among high Japanese naval circles 
favored the latter alternative. Probably 
wishful thinking entered the picture here, 
for it was obviously to the advantage of 
the Japanese to concentrate their naval 
forces in the more southerly waters. The 
1st Mobile Fleet, commanded by Admiral 
Jisaburo Ozawa, was soon to be moved to 
Tawi Tawi in the Sulu Archipelago, and 
it was to this force that major responsibility 
for carrying out Operation A-GO was as- 
signed. The Japanese were already suffer- 
ing a shortage of both fuel and tankers, 
and should Ozawa extend the range of his 
operations as far north as the Marianas he 
would take considerable logistical risks. 2 

Before A-GO could be executed, an 
American thrust in another quarter caused 
Toyoda to change his plans. On 27 May 
General MacArthur's forces invaded the 
island of Biak in the Geelvink Bay area of 
New Guinea and placed the Japanese ad- 
miral in a dilemma. If Biak were lost to 
the invaders, the success of A-GO could 

easily be jeopardized by American aircraft 
based on that island. On the other hand, 
to reinforce Biak would entail at least a 
temporary dispersion of forces, and of 
course the first principle upon which A- 
GO was based was that of concentration 
of force. Faced with this choice, the Jap- 
anese decided to accept the risks of 
dispersion and to dispatch some of their 
ships and planes to the Biak area in an 
effort to drive the Americans off. This de- 
cision was reflected in a new plan of 
operations known as KON. H 

Three times within eleven days Japanese 
naval forces sailed forth for Biak with troop 
reinforcements. The first of the expeditions 
turned back on 3 June, having lost the 
element of surprise on being sighted by 
American submarines and planes. The sec- 
ond was struck by B-25's and suffered one 
destroyer sunk and three others damaged 
before the entire task force was chased 
away by American warships. The third, 
which included the superbattleships Yama- 
to and Musashi, the light cruiser Noshiro, 
and six destroyers, all detached from the 
1st Mobile Fleet, was abruptly called off 
on 1 2 June when Admiral Toyoda received 
definite word that Admiral Spruance's 
forces were attacking the Marianas. 

Thus, the American soldiers on Biak 
were saved from further naval harassment 
by the timely appearance of Central Pacific 
forces off the Marianas. By the same token, 
the invasion of Biak by Southwest Pacific 
forces was to prove a boon to Admiral 

2 Information on Operation A-GO is derived 
from: Japanese Studies in World War II, 60 and 
97; USSBS, Campaigns, pp. 213-72; USSBS, In- 
terrogations, II, 3 r-6. 

3 Information concerning KON and its effects 
is derived from: Smith, Approach to the Philip- 
pines, Ch. XV; Samuel Eliot Morison, History of 
United States Naval Operations in World War II, 
Vol. VIII, New Guinea and the Marianas, March 
1944-August 1944 (Boston, Little, Brown and 
Company, 1953), PP- 1 19-33- 



Spruancc. Japanese plans for A-GO relied 
heavily on the support of naval land-based 
planes of the ist Air Fleet stationed in the 
Marianas, Carolines, and Palaus, but one 
third to one half of all these planes were 
sent to Sorong and other bases in western 
New Guinea in response to the invasion of 
Biak. There, large numbers of pilots fell 
prey to malaria, and most of the aircraft 
were lost either to U.S. action or to bad 
weather. By the time the U.S. Fifth Fleet 
showed up to meet the challenge of A-GO, 
the land-based aircraft available to Toyoda 
had been sizably reduced in number. 

On 1 1 June the Japanese admiral re- 
ceived word of Mitscher's carrier strike 
against Saipan and immediately suspended 
the KON operation, ordering the task 
force bound for Biak to join forces with 
the main body of Ozawa's i at Mobile Fleet. 
Ozawa himself sorticd from Tawi Tawi two 
days later, and on the morning of the 1 5th 
Operation A-GO was activated. Contrary 
to earlier Japanese expectations, the Amer- 
icans had chosen to attack the Marianas 
rather than the western Carolines. Hence 
the scene of the impending "decisive fleet 
engagement" could only lie somewhere in 
the Philippine Sea — that vast stretch of 
ocean between the Philippines and the 
Marianas. 4 

4 The following account of the Battle of 
the Philippine Sea is derived from CINCPAC- 
CINCPOA Opns in POA— Juii 44, Annex A, Part 
VII; Morison, New Guinea and the Marianas, 
Chs. XIV-XVI. 

The Philippine Sea was so named as the result 
of a recommendation made by Admiral Nimitz 
in 1944, at the time of operations against the 
Marianas. The name was officially approved by 
the U.S. Board on Geographic Names in March 
1945. The Philippine Sea applies to that area 
limited on the north by Japan, on the east by the 
Bonins and the Marianas, on the south by the 
Carolines, and on the west by the Philippines, 
Formosa, and the Ryukyu Islands. 

On the evening of 15 June Ozawa's fleet 
had completed its progress from Tawi Tawi 
up the Visayan Sea and through San Ber- 
nadino Strait into the Philippine Sea. On 
the next afternoon it was joined by the 
KON force that had been diverted from 
Biak. Both fleets were sighted by American 
submarines, and it was apparent that the 
Japanese were heading in a northeasterly 
direction toward the Marianas. 

All together, Ozawa had mustered 5 car- 
riers, 4 light carriers, 5 battleships, 1 1 
heavy cruisers, 2 light cruisers, 28 destroy- 
ers, and 430 carrier-based combat aircraft. 
He was outnumbered by the Americans in 
every respect except in heavy cruisers. 
Spruancc had at his disposal 7 carriers, 8 
light carriers, 7 battleships, 8 heavy cruis- 
ers, 13 light cruisers, 69 destroyers, and 
891 carrier-based planes. 5 The mammoth 
American fleet was divided into four car- 
rier task groups under Admiral Mitscher, 
Commander, Task Force 58. Mitscher was 
in tactical command, but his major tactical 
decisions had to be approved by Spruancc 
as Commander, Fifth Fleet. 

By the morning of 18 June all four 
American carrier groups had rendezvoused 
and were steaming in a southwesterly di- 
rection toward the approaching enemy. 
Spruance had ordered: 'Action against the 
enemy must be pushed vigorously by all 
hands to ensure complete destruction of his 
fleet," but had added the precautionary 
note, "Task Force 58 must cover Saipan 
and our forces engaged in that opera- 
tion.'""' That night Admiral Mitscher 
learned the full meaning of this qualifica- 
tion when his superior ordered him to 
change course to the east and maintain it 

r> Figures are from Morison, Neiv Guinea and 
the Marianas, p. 233. 
" Ibid., pp. 243, 250. 



until daylight. Mitscher protested but was 
overruled. Admiral Spruance was fearful 
that Ozawa might attempt an end run un- 
der cover of darkness and put the Japanese 
fleet between him and Saipan. The Fifth 
Fleet commander was unwilling to jeopar- 
dize the landing operations even if it meant 
a delay in closing with the enemy fleet. 
Actually, no such end run was contem- 
plated by the Japanese commander, but 
Spruance had no way of knowing that at 
the time. 

On the morning of the 19th, after the 
American carriers had turned west again, 
Ozawa's planes, which were lighter and less 
well armed and therefore capable of greater 
range than their American rivals, delivered 
the first blow. In four separate raids lasting 
for almost five hours Japanese planes 
roared over the horizon in a futile effort to 
knock out Mitscher's mighty fleet. Out of 
all the American surface vessels present, 
only one was hit — the battleship South Da- 
kota, which lost 27 men killed and 23 
wounded, but was not seriously damaged. 
For the rest, the raids were broken up and 
the raiders destroyed or turned back by 
the combined might of American ships' fire 
and planes, chiefly the latter. Later that 
afternoon American strikes on Guam and 
Rota, which had been ordered for the 
morning, were resumed. By evening the 
"Great Marianas Turkey Shoot" was over 
with disastrous results to the Japanese. Out 
of 430 carrier planes, Ozawa lost 330. 
Some went down under the fire of Amer- 
ican ships and planes; others were de- 
stroyed on Guam and Rota; and still 
others were counted as operational casual- 
ties. Against this, only twenty-four Ameri- 
can planes were shot down and six lost 
operationally. The same day, two Japanese 
carriers, Shokaku and Taiho (Admiral 

Ozawa's flagship) were sunk by American 
submarines operating well to the south of 
Mitscher's fleet. 

That night Ozawa changed course to the 
northwest hoping to put distance between 
himself and the American fleet and to al- 
low himself opportunity to refuel. Mitscher 
held to a westerly course in the belief that 
it would bring him across the track of his 
enemy. However, he could not send out 
night air patrols because none of his carrier 
aircraft were equipped with search radar, 7 
and not until late the following afternoon 
was aerial contact finally made with the 
Japanese fleet, which was now heading in 
the general direction of Okinawa, Mitscher 
immediately launched a twilight air attack 
that succeeded in destroying about 65 of 
Ozawa's remaining 100 aircraft, sinking 
the carrier Hiyo, hitting another carrier 
and a battleship, and damaging two fleet 
oilers to the extent that they had to be 
scuttled. American plane losses came to 1 00, 
mostly incurred through crashes when the 
returning planes tried to land on their car- 
riers after dark. Personnel casualties were 
not so heavy, coming to only 49. 

Thus ended the Battle of the Philippine 
Sea. Mitscher would have detached his 
battleships, cruisers, and destroyers to pur- 
sue and destroy the fleeing enemy, but 
Spruance refused to break up the fleet. It 
would have made no difference anyway 
since Ozawa was by now too far away to 
be overhauled. 

Despite the escape of six carriers and 
their escorts, the Imperial Navy had suf- 
fered a severe blow — one from which it 
never recovered. In the opinion of Samuel 
Eliot Morison, the Battle of the Philippine 

7 Ltr, Admiral Spruance to Maj Gen A. C. 
Smith, 28 Feb 55 , OCMH. 



Sea "decided the Marianas campaign by 
giving the United States Navy command 
of the surrounding waters and air. Thus, 
the Japanese land forces on Saipan, Tinian, 
and Guam were doomed, no matter how 
bravely and doggedly they fought." 8 

There can be no doubt of the decisive 
influence of the sea battle on the ultimate 
outcome of the land campaigns in the Mar- 
ianas. On the other hand, the immediate 
effects were not altogether beneficial from 
the point of view of the troops fighting 
ashore on Saipan. On first getting word of 
the approach of the Japanese Fleet, Ad- 
miral Spruance had detached from Admiral 
Turner's attack force five heavy cruisers, 
three light cruisers, and twenty-one de- 
stroyers to supplement Task Force 58,'' 
This left Turner without adequate fire sup- 
port for his transport shipping, which was 
still in the process of unloading at Saipan, 
Consequently, mast of the transports re- 
tired well to the eastward of Saipan on the 
night of 17 June and remained away from 
the Saipan area until the Battle of the Phil- 
ippine Sea was over. 10 The withdrawal of 
these transports naturally interrupted un- 
loading and imposed additional strains on 
the already overburdened logistical pro- 
gram on Saipan. 


No aspect of an amphibious landing 
against a hostile shore presents more com- 
plex problems than that of transporting 
supplies from ship to shore and allocating 
them at the proper time and place and in 
the proper amounts to the troops that need 


Morison, New Guinea and the Marianas, p. 

9 Ibid., p, 242. 

10 TF 5a Opns Rpt Marianas, Inc.l A, pp. 6-7. 

them. Similarly, no phase of an amphibious 
operation is so likely to become disorgan- 
ized and even disorderly. Ordinarily, the 
assault landing craft and vehicles move 
from ship to shore in scheduled wave for- 
mations and in a fairly methodical fashion. 
Once ashore the troops deploy and eventu- 
ally move inland according to prearranged 
plan. Supplies, on the other hand, cannot 
move off the beach under their own power. 
More often than not they are dumped at 
the water's edge in a haphazard fashion 
by landing craft whose naval crews are pri- 
marily interested in putting out to sea 
again. The supplies stay at the shore line 
until shore parties can segregate them in 
some order on the beaches or until mechan- 
ical transportation comes ashore to haul 
them in to inland dumps. To the casual 
observer at least, the pile up and congestion 
of supplies at the shore line during the first 
phase of a normal amphibious assault pre- 
sents a picture of total chaos. 

To be sure, in a well-conducted amphib- 
ious operation the chaos ls often more ap- 
parent than real, but even under the best 
conditions the problem of ship-to-shore 
supply is a complicated one and not easy 
of solution. At Saipan it was further com- 
plicated by local circumstances, which 
were formidable, although not unique. On 
the first day unloading was hampered by 
heavy artillery and mortar fire on the 
beaches that did not cease altogether until 
three days later. Hydrographic conditions 
were unfavorable to a steady movement of 
supplies and equipment in to the beaches. 
The uncertain naval situation made it 
necessary for the transports to retire to sea- 
ward each of the first three nights. Finally, 
for the next five days and nights most of 
the transports stayed at sea awaiting the 



outcome of the Battle of the Philippine 
Sea. 11 

Enemy harassment of the beaches and 
the unfavorable hydrographic conditions 
offshore were of course felt most seriously 
by the two Marine divisions during the first 
two days of the operation. Sporadically, 
enemy fire caused all unloading work to be 
suspended as shore parties took cover. The 
beaches on the flanks of the landing area 
were completely inaccessible to boats of 
any kind. LST's and LCT's could ground 
on the abutting reef, but supplies from that 
point to shore had either to be man- 
handled or transferred to LVT's and 
DUKW's. Some landing craft could reach 
shore at the interior beaches by way of the 
narrow channel off Charan Kanoa, but at 
low tide its use was restricted to those of 
the most shallow draft, and at all times it 
was congested because both assault divi- 
sions were using it.' 2 

Inevitably, too, along the six thousand 
yards of beach there was some mix-up of 
supplies in spite of the elaborate organiza- 
tion to supervise the unloading of the 
transports and the movement of supplies 
to the troop units to which they were al- 
located. In accordance with standard am- 
phibious doctrine, this task was shared by 
naval beach parties and ground force shore 
parties. The beach parties supervised the 
unloading of the transports and the pro- 
gress of landing craft and vehicles to the 
shore line. Also, they marked channels and 
controlled traffic in the lagoon. In com- 
mand of these operations was a force mas- 
ter who had under him two transport 
group beachmasters, one for each Marine 

division, each of whom in turn commanded 
two transport division beachmasters, one 
for each assault regiment. All of these naval 
officers were landed as soon as satisfactory 
lateral communications had been estab- 
lished, and each was provided with a 
communication team of one officer, five 
radiomen, and five signalmen. 13 

Paired with the naval beachmasters and 
working in close co-ordination with them 
were the Marine and Army shore party 
commanders whose job it was to control 
traffic on the beaches themselves, receive 
the supplies as they were landed, and dis- 
tribute them to the appropriate troop 
units. Each Marine division was author- 
ized a shore party of 98 officers and 2,781 
enlisted men. The 2d Marine Division 
based its shore party organization on the 
pioneer battalion of its engineer regiment. 
Nine teams were organized under three 
shore party group headquarters. The or- 
ganization of the 4th Marine Division's 
shore party differed somewhat in that two 
shore party groups were set up, each with 
three teams. These were drawn from per- 
sonnel of the 121st Naval Construction 
Battalion as well as from the pioneer bat- 
talion of the division engineering regi- 
ment. 14 Each of the three Army regiments 
had its own shore party battalion — the 
i52d Engineers for the 165th Infantry, the 
34th Engineers for the 105th, and the 
1341st Engineers of the 1165th Engineer 
Group for the 106th Infantry. 15 

Notwithstanding this system of inter- 
locking and parallel controls, the first two 
days of the operation frequently saw sup- 
plies of the 2d Marine Division being 

11 COMINCH P-007, Ch. 5, p. 7. 

13 4th Marine Div Opns Rpt Saipan, Annex D, 
pp. 8-9; 2d Marine Div SAR. Phase I. Forager, 
Part II, p. 32. 

13 TF 51 Opns Rpt Marianas, Ind C, pp. 5-6. 
14 NTLF Rpt Marianas, Phase I, Incl J, Engr 
and Shore Party Rpt, p. 2. 

is 27th Inf Div G-4 Rpt, pp. 12, 17. 



Shore Parties Unloading Supplies on Blue Beach i 

dumped on the beaches of the 4th Division 
and vice versa. When the 27th Division 
began to land the situation rapidly deter- 
iorated. On the night of 16 June the 165th 
Infantry commenced to come ashore over 
Blue Beach 1 in the zone of the 4th Marine 
Division. Next day the 105th Infantry 
followed. Few, if any, preliminary plans 
had been made by higher headquarters to 
cover the details of landing the reserve di- 
vision, and the division itself, in its initial 
planning, had made no provision for a 
landing in this particular area. 16 Hence, 
the process of getting the troops and sup- 
plies ashore inevitably became a makeshift 
proposition. Col. Charles Ferris, Division 
G-4, set up an on-the-spot system of con- 
trols, but was unable to persuade either 
the Navy beachmaster or the senior shore 
party commander at corps headquarters to 
agree to routing cargo to any single beach 

18 See above, p. 41. 

or to order the transports' boats to report 
to a single control craft in the channel so 
that an accurate record could be kept of 
items discharged. The result was that the 
27th Division's supplies were landed over 
several beaches, and the Army troops had 
to scramble and forage to get what they 
needed. 17 

On 18 June all ships carrying troops and 
supplies of the 27 th Division retired east- 
ward of the island to await the outcome 
of the Battle of the Philippine Sea. To add 
to the division's difficulties, the i52d En- 
gineers, which had been assigned as the 
shore party for the 165th Infantry, was 
detached and assigned to corps, leaving 
only the 34th Engineers to perform shore 
party functions for the two Army regi- 
ments that landed. By the morning of the 
19th the supply situation within the di- 
vision had become critical. True, there was 

17 27th Inf Div G-4 Rpt, p. 17. 



Caterpillar Tractor Pulling Ammunition Pallets From an LCM 

enough food and water on hand for 
immediate needs, but only by dint of bor- 
rowing K rations from Marine dumps and 
capturing the water cisterns on Aslito field. 
The quantities of Class II, III, and IV 
(organization equipment, fuel and lubri- 
cants, miscellaneous equipment) supplies 
on hand were almost negligible. Small arms 
ammunition would last four days, but 
there were only about 600 rounds of 155- 
mm. ammunition available for each bat- 
talion and 1,200 rounds per battalion of 
105-mm. ammunition, most of the latter 
borrowed from the Marines. Of the divi- 
sion's vehicles, there were on shore only 
three 2- 1/2 -ton cargo trucks, twenty- 
three 3/4-ton weapon carriers, and forty- 
nine DUKW's. Not until 20 June did the 
ships carrying the troops of the 106th 

Regiment return to Saipan, and not until 
the 27th were the division's supplies and 
equipment fully unloaded. 18 

The 27th Division was not alone in 
suffering an interruption to the flow of its 
supplies and equipment because of the 
Battle of the Philippine Sea. The hasty 
withdrawal of the naval transports on the 
18th made an orderly discharge of any car- 
go over the proper beaches impossible. 
Priorities were assigned to rations, ammu- 
nition, and fuels, and other items had to 
be neglected. Moreover, in order to dis- 
patch the cargo ships with all possible 
speed out of the danger area, it was neces- 
sary to permit them to unload over the 
beach that was the handiest. Thus, it was 

Ibid,, pp. 18-19. 



Pontoon Gauskway and Barge in Ciiaran Kanoa Harbor 

impossible to prevent a division's sup- 
plies from being scattered among all the 
dumps. 10 The only advantage enjoyed by 
the two Marine divisions in this respect lay 
in the three full days they had had to 
unload their cargo before the general exo- 
dus of naval shipping, but this was at least 
partly offset by the fact that for most of 
the time their landing beaches were under 

Meanwhile, during the period when most 
of the transports were cruising east of Sai- 
pan, shore parties were furiously at work 
improving the beach approaches and elim- 
inating obstacles to a more rapid delivery 
of supplies across the reef. Twelve ponton 

sections had been hauled to Saipan lashed 
to the sides of LSTs, and others came 
later, side-carried by ships of the first gar- 
rison echelon. By 18 June naval Seabccs 
had floated three of these and commenced 
construction of a causeway pier off Charan 
Kanoa, which was increased in length as 
fast as additional sections could be ob- 
tained from LST's on their return from 
retirement to sea. 2 " Next day the 34th En- 
gineer Battalion opened up Yellow Beach 
3 just north of Agingan Point by blowing 
two channels through the reef to permit 
small boats to discharge on the shore dur- 
ing high tide. The battalion also rigged up 
a crane on an overhanging point off the 
beach to enable it to unload materiel di- 

,9 NTLF Rpt Marianas, Phase I, Incl F, G-4 
Rpt, Part II, p. 5; Incl J, p. 5. 

COMINCH P-007, Ch. 5) p. 15. 



rectly from landing craft into trucks 
waiting on a level above the beach itself. On 
21 June the 1341st Engineer Battalion re- 
moved all the mines from White Beach 1 
below Agingan Point and prepared access 
roads from the beach. Naval underwater 
demolition teams searched for antjboat 
mines and blew landing slips in the off- 
shore reef for LST's and LCT's. 21 

One factor that eased the 27th Division's 
unloading problems was that a large per- 
centage of its supplies had been palletized 
before embarkation from Oahu. Unlike the 
two Marine divisions, which were skeptical 
of the process, the 27th Division had re- 
sponded enthusiastically to General Hol- 
land Smith's administrative order that 25 
to 50 percent of all supplies and two to five 
units of fire be palletized. In fact, the Army 
division had palletized almost 90 percent of 
all its supplies and had reason to be grate- 
ful for its own forehandedncss. 22 Securing 
the materiel to wooden pallets permitted 
a more rapid unloading of landing craft at 
the beaches, released working parties that 
otherwise would have been engaged in the 
arduous labor of transferring cargo from 
landing craft into trucks, and reduced the 
number of men at the landing beaches in 
positions exposed to enemy fire. In the 
opinion of Holland Smith, "These advan- 
tages were clearly manifest at Saipan when 
palletized supplies of the 27th Division 
were handled as against unpalletized 
supplies of the 2d and 4th Marine Di- 
visions." L ' y 

Not all of the logistical difficulties that 
beset the fighting troops were due to un- 
loading difficulties. Some shortages can be 

21 27th Inf Div G-4 Rpt, p. 19; NTLF Rpt 
Marianas, Phase I, Incl J, p. 5. 

22 See above, pp. 48-49. 

23 Quoted in GOMINCH P-007, Ch. 5, p. n. 

traced back to the point of embarkation 
and are attributable to insufficient ship- 
ping. This was particularly true in the case 
of motor transportation. There was simply 
not enough space aboard the transports 
assigned to the operation to stow all of the 
vehicles of the three infantry divisions and 
of the XXIV Corps Artillery. General 
Smith's headquarters cut the table of or- 
ganization and equipment allowances, and 
his own allowances were reduced again be- 
cause of inadequate shipping space. In the 
end, out of all motor transport vehicles al- 
lowed by corps, only 94 percent of the 
ambulances, 83 percent of the trucks, 71 
percent of the trailers, and 75 percent of 
the tractors could be embarked from 
Oahu. 24 Although these cuts were distri- 
buted more or less equally among all the 
units involved, the 27th Division suffered 
somewhat less than the others, being able 
to carry with it 86 percent of its trucks, 
99 percent of its trailers, and 99 percent of 
its tractors. 25 However, this advantage was 
more than offset when, after arrival on 
Saipan, corps headquarters commandeered 
thirty-three of the Army division's 2-1/2- 
ton trucks and refused to return them even 
as late as 6 July. 20 ' 

As a partial compensation for the short- 
ages in standard types of motor transporta- 
tion, a substantial number of DUKW's 
was provided for the Saipan operation- ■ 
more than had hitherto been used in the 
Central Pacific. All together, 185 of these 

- 4 NTLF Rpt Marianas, Phase I, Incl F, G-4; 
Incl A, Annex A, p. I. 

25 Ibid. The following tabic gives the distribu^ 
tion of motor transportation among the three 

Division Trucks Trailer; Tractors 

id Marine 437 222 16 

4th Marine 558 101 40 

27th Infantry 624 196 121 

26 27th Inf Div G-4 Rpt, pp. 20-23. 



vehicles were embarked. Each infantry di- 
vision had a DUKW company attached, as 
did XXIV Corps Artillery. 27 The DUKW's 
initial function was to land the artillery. 
This had entailed some modification both 
of the DUKW's bodies and of the 105-mm. 
howitzer wheels. After the landing phase, 
the amphibian trucks were used continual- 
ly throughout the campaign, chiefly for 
hauling ammunition from shipboard or 
supply dumps to artillery emplacements 
and as prime movers for 105-mm. howit- 
zers. In the opinion of General Holland 
Smith's G-4 officer, Colonel Anderson, 
GSC, "the DUKW was the outstanding 
single type of equipment employed in this 
operation." 2H 

Later, as the fight in the central and 
northern part of Saipan progressed, one 
other serious supply shortage manifested 
itself. The heavy demand for artillery sup- 
port and close-in infantry support by 81- 
mm. and 60-mm. mortars created an un- 
expected drain on the mortar ammunition 
supply. Previous experience in the Central 
Pacific had seemed to indicate that a total 
of seven units of fire would be sufficient 
for Saipan, but this proved to be too low 
an estimate, and on the basis of experience 
in the Marianas ten units was recom- 
mended for future operations. 20 The initial 
fault in not loading enough ammunition 
aboard the assault ships was compounded 
by the fact that resupply ships were fre- 
quently not vertically loaded, thus making 

21 1 st Amphibian Truck Company for the 2d 
Marine Division ; 2d Amphibian Truck Company 
for the 4th Marine Division ; Provisional DUKW 
Company (from Quartermaster Company) for the 
27th Division; 477th Amphibian Truck Company 
for XXIV Corps Artillery. 

2H NTLF Rpt Marianas, Phase I, Incl F; Incl 
B, p. 1. 

-'-' For the Pacific Ocean Areas unit of fire, sec 
Appendix B. 

it difficult for the troops to get their am- 
munition ashore when they needed it. Also, 
the first ammunition resupply ship was late 
in arriving in the area, and many of 
these vessels withdrew at night because of 
threatened air attacks, thus making a 
build-up of reserves impossible. 30 

The extent to which these shortages and 
delays, avoidable and unavoidable, affected 
the course of the battle on Saipan cannot 
definitely be determined. The shortages 
and delays were real enough, but the fact 
that they were reported in such detail by 
supply officers and others might be taken 
to be as much in evidence of the wealth of 
materiel to which Americans in combat 
were accustomed as of any real privation 
suffered on Saipan. The records show no 
single instance wherein any infantry or ar- 
tillery unit had to cease fire for want of 
ammunition or became completely immo- 
bilized for lack of transportation. On the 
other hand, it can be assumed that any 
defects in a supply system automatically 
impede the progress of ground troops. It is 
highly probable that had more supplies 
been on hand and had they reached the 
front lines in a more expeditious fashion, 
the combat troops would have been able 
to move against the enemy with greater 
force and speed, 

Postlanding Naval Gunfire Support 

On Saipan, as elsewhere in the island 
warfare typical of the Pacific, one of the 
most effective weapons in support of the 
infantry proved to be ships' fire. Naval ves- 
sels ranging in size from LCI gunboats to 
old battleships, and mounting guns of cal- 
ibers from 20-mm. to 14 inches, cruised the 

!i0 NTLF Rpt Marianas, Phase I, Incl F, Part 
II, p. 7. 



coasts of the island prepared at all times 
to support the troops on call, lay down 
preparatory fire, illuminate the night with 
star shells, and perform a host of other 
duties. True, the configuration and terrain 
of Saipan imposed some natural limitations 
on the fullest exploitation of this support. 
Naval guns have a flat trajectory and the 
mountains and hills of the volcanic island 
often masked fire from the sea. Also, the 
reefs fringing many parts of the island kept 
the larger vessels from approaching within 
optimum range for direct fire at some tar- 
gets. On the other hand, many of the 
enemy's guns and installations were em- 
placed in defilade in valleys that ran 
perpendicular to the shore line on the east 
and west coasts, and against these naval 
gunfire could be particularly effective. 
The caves along the shore line offered ideal 
hiding places for enemy troops and were 
also ideal targets for ships firing from the 
sea, especially for the vessels of more shal- 
low draft. 31 

In general, two types of controls were 
set up to permit the co-ordination between 
troops and ships so necessary to efficient 
operations of this sort. 32 For close support 
missions it was customary each day to as- 
sign a certain number of vessels to each 
infantry battalion in the assault- — usually 
two or three destroyers. During the entire 
course of the operation 2 old battleships, 
2 heavy cruisers, 3 light cruisers, and 39 
destroyers delivered call fires at various 
times. Attached to each battalion was a 
shore fire control party consisting of naval 
and ground force personnel furnished by 
the 1 st, 2d, and 295th Joint Assault Signal 

31 TF 52 Opus Rpt Marianas, Incl F, p. 13. 

3 2 This account of naval gunfire is derived 
chiefly from NTLF Rpt Marianas, Phase I, Incl 
I, Sec. 3, and TF 52 Opns Rpt Marianas, Incl F. 

Companies, which were attached, respec- 
tively, to the 4th Marine Division, the 2d 
Marine Division, and the 27th Infantry Di- 
vision. These parties were in direct radio 
communication with their supporting ships 
and from their positions on shore would 
request fire missions and spot the results. 

This system of control was generally em- 
ployed for two types of missions. First were 
the close support missions fired sometimes 
within fifty yards of friendly troops. De- 
stroyers' five-inch guns were usually used 
for this purpose, and the bulk of all five- 
inch ammunition (139,691 rounds) was 
consumed in this fashion. How effective it 
was is doubtful. Admiral Turner's final 
opinion was, "Field Artillery is much better 
qualified for this type of fire by reason of 
its greater accuracy and smaller burst 
patterns." 33 The second type of mission 
usually controlled by shore fire control par- 
ties was night illumination. Star shells 
were fired on request of the infantrymen 
to prevent infiltration, to help stop coun- 
terattacks, and to keep enemy activity to 
their immediate front under surveillance. 
Unfortunately, there were not enough of 
these projectiles on hand to satisfy the 
wants of the troops, and after the first 
night a quota of six per hour per ship had 
to be imposed except during emergencies. 

Except for these two types of missions, 
request for all other sorts of ships' fire on 
ground targets originated from the naval 
gunfire officer of the Northern Troops and 
Landing Force. His headquarters was set 
up ashore near those of the corps air officer 
and the corps artillery officer and the three 
worked in close co-ordination so as to avoid 
duplication of effort and waste of am- 

TF 53 Opns Rpt Marianas, Incl F, p. 13. 



Under the naval gunfire officer's super- 
vision, all deep support fire missions were 
arranged, including preparation fires, de- 
liberate and methodical destruction fires, 
counterbattery, harassing, and interdiction 
fires, and fires on targets of opportunity. 
Since these missions were not controlled by 
shore fire control parties, it was considered 
necessary to fix definite safety limits and 
to specify safe lines of fire. Preparation fires 
were not brought closer than 1,500 yards 
from the nearest friendly troops. Deliberate 
and moderate destruction fires, fires on tar- 
gets of opportunity, and counterbattery, 
harassing, and interdiction fires were usual- 
ly confined to areas 2,500 yards from the 
front line. On the whole these various 
missions were executed far more effectively 
than were close support fires, and it was in 
this field that naval gunfire won its laurels 
at Saipan. Other chores ably performed by 
the support ships were the. guarding of 
Saipan against amphibious reinforcements 
from Tinian, neutralization of the airfields 
at Marpi Point and Ushi Point, Tinian, 
and destruction of enemy cave positions 
along the seacoast that were inaccessible, 
to anything but the 40-mm. fire of LCI 

The testimony of prisoners of war cap- 
tured on Saipan leaves no doubt of the 
impression made on the Japanese by Amer- 
ican naval gunfire. Maj. Takashi Hira- 
gushi, 43d Division intelligence officer, 34 
testified, "the most feared of . . . [Amer- 
ican] weapons was the naval shelling 
which managed to reach the obscure 
mountain caves where . . . CP's were lo- 
cated." 35 A captured Japanese lieutenant 

S4 He was mistakenly identified as Maj, Kiyoshi 
Yoshida, 31st Army intelligence officer. 

:lH NTLF Rpt Marianas, Phase I, Incl I, Sec. 3, 
P- >9- 

declared that the greatest single factor in 
the American success was naval gunfire. 
When asked how he distinguished between 
naval gunfire and land-based artillery, he 
laughed and said that it was not difficult 
when one was on the receiving end. Every- 
one in the hills "holed up" and waited when 
a man-of-war started to fire. 38 Other Jap- 
anese prisoners of war, when interrogated 
on the matter, were in almost unanimous 
agreement. Perhaps the highest testimonial 
of the efficacy of this particular weapon 
came from General Saito himself when he 
wrote on 27 June, "If there just were no 
naval gunfire, we feel with determination 
that wc could fight it out with the enemy 
in a decisive battle." 37 

Close Air Support 

Once the assault troops had landed on 
Saipan and established their beachhead, 
the role of aircraft for the remainder of the 
operation was twofold. First, and most im- 
portant, it was to keep the battlefield iso- 
lated from the inroads of enemy air and 
surface craft. Second, it was to support the 
advance of the ground troops in somewhat 
the same manner as naval gunfire and ar- 

After the Battle of the Philippine Sea no 
serious threat of enemy air intervention re- 
mained, and except for occasional nuisance 
raids the troops on Saipan could enjoy 
virtual immunity from that quarter. There- 
after, the planes of Mitschcr's Task Force 
58 were employed on occasional troop 
support missions, while Admiral Turner's 
escort carriers provided the aircraft for 
combat air patrols and antisubmarine 
patrols, 38 Once Aslito airfield was captured 

:,H Ibid., p. 20. 

:l7 CINGPAC-CINGPOA Item 9983-85, p. 25. 

3R TF 51 Opns Rpt Marianas, Incl I, p. 4. 



and put into operation, these duties were 
shared by P-47's of the 19th and 73d 
Fighter Squadrons, Seventh Air Force. 39 

Whether in deep support or close sup- 
port, the planes assigned to assist the 
ground forces flew three types of missions 
— bombing, rocketing, and strafing. Of 
these, the first was the least effective in 
knocking out comparatively small targets 
such as gun installations. After the initial 
softening up of the landing beaches, bomb- 
ing missions were ordinarily employed 
against enemy troop concentrations, supply 
dumps, and buildings. The first extensive 
use of aircraft rockets in the Central Pacific 
was on Saipan. The rockets proved to be the 
most valuable weapon for support aircraft, 
in spite of the fact that there was insuffi- 
cient training in its use and that no delay 
fuzes were available. The most common 
technique for close support missions was 
strafing, which was not only effective 
against the enemy but safer for friendly 
troops. 40 

Troop requests for close air support 
were radioed by air liaison parties attached 
to each regiment and battalion. The re- 
quests were filtered through division and 
corps headquarters, each of which had the 
opportunity of rejecting them before final 
decision was made by the Commander, 
Support Aircraft, Capt. Richard F. White- 
head, USN, who was aboard Admiral 
Turner's flagship. Once a strike was or- 
dered, it would be controlled either by 
Captain Whitehead himself, by the support 
aircraft commander on Holland Smith's 
staff, by the air co-ordinator (who was a 
group or squadron leader from one of the 
participating carriers and was on station 

over the island at all times during daylight 
hours), or by the flight leader assigned to 
the particular mission. 

Air liaison parties on the ground had no 
direct radio communication with the planes 
and were therefore unable to coach the pi- 
lots into their targets. Targets were desig- 
nated in a variety of ways. Sometimes the 
infantry marked them with white phos- 
phorus mortar shells. At others, planes 
flew dummy runs and waited to execute 
their missions until battalion air liaison par- 
ties notified the Commander, Support 
Aircraft, who in turn notified the flight 
leader if the runs were made on the correct 
area. Fluorescent panels were used to mark 
the front lines of the troops. 41 

The highly centralized system of close air 
support control used at Saipan had the 
advantage of reducing to a minimum the 
danger of duplication of missions and of 
planes bombing and strafing within friend- 
ly lines. On the other hand, it was time 
consuming to a degree that was highly 
unsatisfactory to the troops. The time lag 
between requests for and execution of an 
air strike was sometimes more than an 
hour and seldom less than a half hour. 42 

One reason for the delay was the diffi- 
culty of co-ordinating air with the other 
supporting arms. No single co-ordinating 
agency had been established before the in- 
vasion of Saipan. This created no especially 
difficult problem when it came to co- 
ordinating air and naval gunfire, since by 
mutual agreement naval gunfire was lifted 
in certain areas when requested by the 
Commander, Support Aircraft, and air at- 
tacks were stopped on the request of firing 
ships. On the other hand, the co-ordination 

3a AAF Hist Div, Army Air Forces in the Mari- 
anas Campaign, pp. 15-26, MS, OGMH. 

lu TF 51 Opns Rpt Marianas, Incl I, p. 11. 

41 Ibid., pp. 9-10. 

42 4th Marine Div Opns Rpt Saipan, Incl H, 
PP- 97-103- 



of air and artillery presented a more 
difficult problem because of the higher ord- 
inate of artillery pieces, their rapid rate of 
fire, and the lack of central control for 
the four separate artillery units. 43 For these 
and other reasons, close air support was 
the least satisfactory of the three support- 
ing arms. 


Artillery support was of course pro- 
vided by the three divisions' organic pieces, 
as well as by the twenty-four 155-mm. 
guns and twenty-four 155-mm. howitzers 
of XXIV Corps Artillery, which was com- 
manded by General Harper. Corps artillery 
commenced to land and go into position on 
18 June, and by 22 June all four battalions 
were ashore and firing. 44 The two 1 55-mm. 
howitzer battalions and one of the gun bat- 
talions were emplaced 1,500 to 2,000 yards 
south of Gharan Kanoa on the low, flat, 
plain adjacent to Yellow Beaches, while the 
other gun battalion was emplaced on the 
higher ground just southwest of Aslito air- 
field. Initially, all battalions faced north 
on Saipan except for Battery B, 531st 
Field Artillery Battalion, which was posi- 
tioned to fire on Tinian. 45 On 27 June the 
front lines had advanced to an extent call- 
ing for a forward displacement of the heavy 
battalions of the corps, and by the 28th 
all had been displaced to positions northeast 
of Magicienne Bay. On 7 July the 225th 
Field Artillery Battalion displaced again, 
this time to the northeastern edge of Kag- 
man Peninsula. In addition to supporting 

the troops on Saipan, XXIV Corps Artil- 
lery had the job of guarding the back door 
to Tinian. Observation posts overlooking 
the southern island were manned twenty- 
four hours a day, and various harassing 
and destructive missions were fired on Tin- 
ian airfields and other targets on that island 
throughout the Saipan operation. 4 ' 1 

For the most part corps artillery was as- 
signed the job of delivering deep support 
fires for the advancing troops, and a mini- 
mum safety band of 1,500 yards in front 
of the infantry was established. The divi- 
sion's batteries engaged in night harassing 
fires, preparation fires in advance of the 
daily infantry jump-offs, fires on targets of 
opportunity, and call fires at the request 
of the troops. 47 On several occasions di- 
vision artillery fired rolling barrages. Close 
liaison was maintained between corps and 
each division artillery headquarters by 
liaison officers numbering as many as three 
per division. A similar system was main- 
tained by the divisions themselves. Each 
light artillery battalion had a command 
liaison officer with its supported infantry 
regiment, and usually the Marine and 
Army artillery units exchanged liaison of- 
ficers to co-ordinate fires near division 
boundaries. Primary means of communica- 
tion was by wire, although this was not 
altogether satisfactory because the large 
number of tracked vehicles used on Saipan 
made maintenance of wire lines difficult. As 
a substitute, all corps liaison officers were 
provided with truck-mounted radios. 48 

45 TF 51 Opns Rpt Marianas, p. 10. 

** XXIV Corps Arty Final Rpt on Forager 
Opn, Phases I and III, S-3 Rpt, p. 6. 

16 Hq Central Pacific Base Comd, Target Sai- 
pan: A Story of XXIV Corps Artillery, p. 6. 

4fl XXIV Corps Arty Final Rpt on Forager 
Opn, Phases I and III, S-3 Rpt, pp. 7-8. See 
below, Ch. XIII. 

"" Ibid., passim; 27th Inf Div Arty Rpt For- 
ager Opn, Sec. II. 

48 XXIV Corps Arty Final Rpt on Forager 
Opn, S-3 Rpt, p. 16; 37th Inf Div Arty Rpt 
Forager Opn, Annex A, p. 4. 



North Central Saipan from the crest of Mt. T apotchau to the. northern tip of 
the island. Northern end of Death Valley is visible at right center. 

All together, the four artillery units were 
to fire about 291,500 rounds before the 
end of the battle for Saipan. Of these, 
37,730 can be attributed to the corps ar- 
tillery. 43 In spite of this considerable 
volume of fire, there were certain limiting 
factors to the optimum employment of ar- 
tillery. Chief among these was terrain. 
After the major portion of southern Saipan 
had been secured and the main American 

49 The estimate is derived from Hoffman, Sai- 
pan, p. 250; corps figures arc from XXIV Corps 
Arty Final Rpt on Forager Opn, S-3 Rpt, p. 8. 

attack reoriented to the north, General 
Holland Smith disposed almost all of his 
field artillery to support the northward 
thrust. True, one battery of 155-mm. guns 
(later increased to three) was pointed to 
the south against Tinian, but all the re- 
maining pieces were ordered to direct 
their fire against the Mount Tapotchau- 
Death Valley-Kagman Peninsula line and 

The terrain in this central part of Saipan 
presented several problems to the gunners. 
XXIV Corps Artillery was assigned the 



general mission of deep support, which 
meant that most of its targets were located 
in the northern half of the island. Since 
Mount Tapotchau lay athwart the line of 
sight between ground observers and these 
targets, corps artillerymen had to rely en- 
tirely on air spotters. Six L-4 liaison planes 
were assigned for this purpose, and by the 
end of the operation each of the pilots and 
his accompanying air observer had put in 
approximately a hundred hours in the air 
over enemy territory,'"'" 

A more serious problem faced the ar- 
tillerymen of the three divisions whose 
mission was to fire in close support of the 
advancing troops. In the center of the is- 
land, just cast of Mount Tapotchau, lay 
Death Valley, which ran north and south 
along the axis of the attack. Since most of 
the enemy's guns and mortars in this area 
were sighted into the valley from the hills 

60 Hq Central Pacific Base Command, Target 
Saipan, p, 6. 

and cliffs on either side, they could not 
easily be reached by American artillery fir- 
ing from the south. This was one reason 
for the slow progress made by infantrymen 
up the center corridor of the island. Fur- 
thermore, since the troops on the right and 
left pushed on more rapidly than those in 
the center, the front line became more and 
more bent back in the middle. The un- 
evenncss of the line made the adjustment 
of artillery fire all the more difficult. 

Nevertheless, in spite of the fact that 
Saipan was far from an artilleryman's para- 
disc, the main body of Holland Smith's 
troops during the attack to the north did 
at least have continuous artillery support. 
Not so those troops of the 27th Division 
that were left to clean out the Japanese 
who were holed up on Nafutan Point, the 
southeastern tip of the island. Except for 
tanks, naval gunfire, and, later, antiaircraft 
guns, the infantrymen assigned to this mis- 
sion would have to depend entirely on their 
own weapons. 


The Capture of Nafutan Point 

By the end of 18 June the 4th Marine 
Division had penetrated to Magicienne Bay 
and cut the island of Saipan in two. Gen- 
eral Holland Smith's plans for the next 
phase of the operation called for a change 
of direction of the main attack from east 
to north across the breadth of the island. 
Specifically, this meant that the 2d Marine 
Division would hold and consolidate its 
positions on the extreme left flank south of 
Garapan and would constitute the pivot of 
a wheeling movement. The outer end of 
the wheel's spoke would be the right flank 
of the 4th Marine Division resting on Ma- 
gicienne Bay. When the turn had been 
completed the two divisions would be 
abreast and ready to launch their northerly 
drive against the main enemy defense line, 
which now stretched across the island in a 
southeasterly direction from the outskirts 
of Garapan to Magicienne Bay. 

Meanwhile, Nafutan Point and the ap- 
proaches to it along the south coast of Sai- 
pan remained occupied by Japanese troops 
that had to be cleared out before Aslito 
field could be considered entirely safe from 
counterattack and infiltration. 

Nafutan Point is a short peninsula — a 
southward extension of the east coast of 
Saipan. Dominating most of the peninsula 
is a high cragged ridge running in a north 
-south direction not far inland from the 
east coast. This is Mount Nafutan, whose 
highest point is about 407 feet. Its 

northern and western faces are almost 
sheer cliffs. About 400 yards west of the 
northern part of Nafutan mountain lies a 
ridge about 300 feet in height. Although the 
lowlands in the western portion of the 
peninsula and in the valley between Mount 
Nafutan and Ridge 300 were mostly under 
cultivation, the hilly and mountainous 
areas in the east wer e g enerally covered 
with thick underbrush ] (See Map II. ) | 

Compressed into this area by the ad- 
vance of the American troops was a motley 
crowd of Japanese military personnel mixed 
with civilians. All together, the military 
contingent numbered about 1,050. In- 
cluded were survivors of the 317th Inde- 
pendent Infantry Battalion, 47th Indepen- 
dent Mixed Brigade; naval personnel who 
had manned the coastal defense guns 
located near the southern tip of the penin- 
sula; antiaircraft and service troops that 
had been swept out of Aslito field; and 
probably stragglers from many other units. 
The men were under no single command, 
at least in the strict sense of that word, 
but the highest ranking officer seems to 
have been a Captain Sasaki, who com- 
manded the 317th Independent Infantry 
Battalion, 1 

The job of disposing of these people and 
securing Nafutan Point was initially as- 
signed to the two regiments of the 27th 

1 See below, p. 159. 



Agingon Pt Cope Obion 

Aslito Airfield Mi Nafutan 

Seaward Cliff Line, Nafutan Peninsula 

Infantry Division that were ashore by the 
1 8th of June. Holland Smith's headquar- 
ters assumed that the task could be com- 
pleted in a short time, and that it would be 
little more than a mopping-up operation. 
The assumption proved to be optimistic in 
the extreme. Not until 27 June was the 
southeastern extremity of Saipan com- 
pletely cleared of the recalcitrant, if dis- 
organized, enemy troops holed up in that 

Action of ig June 

Holland Smith's orders for the 19th 
directed the 27th Infantry Division to 
"complete missions aasigned" in the pre- 
vious day's order, which meant in effect 
that the division was to push to the east 
coast of Saipan along its entire front in- 
cluding all of Nafutan Point. Jump-off 

hour was set by division orders at 07 30. 2 
In position along the front line from right 
to left (south to north) were the 3d and 
1st Battalions, 105th Infantry, and the 2d 
and 1st Battalions, 165th. The latter regi- 
ment had on the preceding day almost 
reached Magicienne Bay, but the line of 
the 105th bent back sharply to the west- 
ward to a position on Saipan's south coast 
only 700 yards cast of Cape Obiam, 

On the extreme right, the 3d Battalion, 
105th Regiment, met no opposition to 
speak of. During the day not an enemy 
shot was fired except for a few random 
rounds of artillery that were lobbed into 
the battalion's area from Nafutan Point. 
Nevertheless, the rugged terrain along the 
southern coast made progress difficult, and 

2 NTLF Opn Order 7-44, 19 Jun 44; 37th Inf 
Div G-3 Jnl, 19 Jun 44, Msg 5. 



by nightfall the battalion had advanced 
only about i ,800 yards in its zone. s 

To its left and well ahead, the ist Bat- 
talion, 105th, jumped off at 0730, as 
scheduled, with Company A on the right, 
B on the left. After three hours of unop- 
posed progress, the battalion came up 
against the first of the series of ridges that 
flank Mount Nafutan to the northwest. In 
spite of considerable enemy small arms and 
automatic weapons fire, both companies 
reached the top of the ridge without 
trouble, but as they went over the crest to 
a stretch of level ground with clear fields 
of fire they were pinned down by heavy fire 
from five separate pillboxes to their im- 
mediate front. The pillboxes were located 
near the boundary line between the two 
advance companies, and the company com- 
manders drew their units into a semicircle 
around the area and poured fire into it. 
After an hour and a half of futile effort 
to place shaped charges against the pill- 
boxes, both companies pulled back to a 
line below the ridge out of range of enemy 
fire. 4 

On being informed of the situation, reg- 
iment ordered the ist Battalion to re-form, 
move to the left, and try to outflank the 
enemy by an attack from the north, 
rather than by a frontal assault. As a 
prelude to the attack, naval planes were to 
deliver a fifteen-minute air strike, which 
was to be followed by a half hour's con- 
centrated division artillery fire. 5 

Promptly at 16 10 the battalion jumped 
off and almost immediately ran into 
trouble. B Company, on the left, had to 
climb the ridge some distance back from 

:l Love, Battle for Saipan, p. 108; 105th RCT 
Jnl, 19 Jun 44, Msg 6a. 

i Love, Battle for Saipan, pp. 101-06. 

5 27th Inf Div G-3 Jnl, 19 Jun 44, Msg 48. 

the enemy positions in order to execute 
the flanking movement. Once on top, it 
was to attack south. However, in getting 
onto the ridge, the men were held up by 
an exploding artillery dump and had to 
take a circuitous route. No sooner had they 
reached the top of the ridge than the Jap- 
anese opened fire with dual-purpose guns. 
By this time it was 1730, well on toward 

Company A, meanwhile, had not been 
able even to get into position to attack. Be- 
fore it could swing into line on the right 
flank of Company B, it too came under 
fire from the enemy positions and the 
men jumped for cover. One soldier (Pvt. 
Thomas C. Baker) succeeded in knocking 
out one of the enemy's pillboxes with a 
bazooka, but even so the company made 
no substantial progress. Shortly after 1800, 
Colonel O'Brien halted the attack, and the 
whole battalion retired to the line of depar- 
ture for the night. There, Company C re- 
placed Company B. 6 

To the north, the 165th Infantry was 
faring somewhat better. The previous 
evening the regiment had stopped short of 
the shore line, and its first task was to 
complete its penetration to the sea. Ahead 
was a steep slope that ran down to a line 
of cliffs at the water's edge, there to drop 
fifty to sixty yards straight down to the 
ocean. The slope was a coral formation 
studded with sharp rocks and pocketed 
with holes, deep canyons, crevasses, and 
caves. The whole area was heavily over- 
grown with a tangle of vines, small trees, 
and bushes. The only feasible means of ap- 
proach to the shore line was by way of a 
series of parallel paths running eastward 
through the undergrowth. 7 

Love, Battle for Saipan, pp, 107-08. 
7 Ibid,, pp, 96-97. 



27T1-1 Division Troops at Cliff Edge, Nafutan Peninsula 

The regiment jumped off on schedule at 
0730 with the 2d Battalion on the right 
(south), 1 st on the left (north). 8 Only A 
Company on the extreme left had any seri- 
ous trouble. An advance platoon ran into 
a Japanese machine gun position and was 
fired upon from ambush and held up for 
over two hours. By 1300 lead elements of 
both battalions had picked their way cau- 
tiously to the ocean's edge. 9 The only ap- 
parent enemy opposition remaining in the 
area was in a small pocket along the 
boundary line between the Army regiment 
and the 4th Marine Division. During the 
afternoon appeals were sent out over a 

8 1st Bn 165th RCT Jnl, 19 Jun 44, Msg 7. 

9 165th RCT Jnl, 19 Jun 44, Msg 63. 

public address system in an attempt to per- 
suade this isolated remnant of enemy troops 
to surrender, but the action met with no 
success. Before the troops dug in for the 
night, the 1st Battalion, on the left, was 
relieved by the 3d, which had been in re- 
serve during the day. 10 

At the close of operations on the 19th, 
two battalions of the 165th Infantry were 
drawn up in defensive positions along the 
southern coast of Magicienne Bay. The 1st 
and 2d Battalions had completed the pro- 
cess of cutting off the enemy on Nafutan 
Point from the rest of the island. However, 
the leftward swing of the 1st Battalion, 

10 1st Bn 165th RCT Jnl, 19 Jun 44, Msgs 33, 

47, 64. 



Infantryman at the Base of Cliff, Nafutan Peninsula 

105th Infantry, and the slow advance of 
the 3d Battalion, 105th, along the southern 
shore, had resulted in a large gap in the 
middle of that regiment's line. In order to 
fill the gap and protect Aslito airfield, 
Colonel Kelley ordered the 1st Battalion, 
165th, to move back to the airfield and dig 


Along the 27th Division front the night 
was quiet except for two widely separated 
actions. On the south coast, a group of 
twenty to thirty civilians stumbled into the 
perimeter of Company L, 105th Infantry, 
and were all killed. In the Magicienne Bay 
area, about an hour after dark, some twen- 
ty Japanese launched a counterattack 

11 Ibid., Msg62. 

against the right flank of B Company, 
1 65th Infantry, but the attack was broken 
up within half an hour. 12 

Action of 20 June 

The morning of 20 June brought about 
a change in the 27th Division's plans and 
a reorientation of the attack against Nafu- 
tan Point. General Ralph Smith, after re- 
viewing the difficulties encountered the 
preceding day by the 1st Battalion, 105th 
Infantry, in its attempt to assault Nafutan 
Ridge frontally from the west, decided that 
the direction of the attack should be 
changed from eastward to southward. He 

1 3 Love, Battle for Saipan, p. 1 10. 



attached the ist Battalion, 105th, to the 
165th Infantry and then at 0800 called a 
conference of the unit commanders most 
concerned with the new plan of attack. In 
attendance, besides General Smith, were 
his operations officer, Lt. Col. Frederic H. 
Sheldon; Colonel Kelley, commanding of- 
ficer of the 165th Infantry; and the three 
battalion commanders of that regiment, as 
well as Colonel O'Brien who commanded 
the 1 st Battalion, 105th. 13 

As a result of the conference, General 
Smith issued his Field Message No. 1, 
which called for a co-ordinated attack by 
the 165th Infantry, with the ist Battalion, 
105th, attached, southward along the main 
axis of Nafutan Point, to commence at 
1000. The day's objective was a line drawn 
across the peninsula about halfway be- 
tween the line of departure and the 
southern tip. 14 The 3d Battalion, 105th, 
in the meanwhile would continue to ad- 
vance eastward along the southern coast 
until it could close lines with the rest of 
the division in a tightening noose around 
Nafutan. ln 

For the main attack down the peninsula, 
the line-up of units from right to left (west 
to east) was: Companies C and A of the 
105th Infantry and Companies I, K, F, 
and G of the 165th. The terrain to the 
front of the three battalions varied. Im- 
mediately ahead of the right flank of the 
ist Battalion, 105th, the ground was fairly 
smooth with no serious obstacles. On its 
left the land sloped upward abruptly to 
a cane-covered plain in Company I's zone. 
Between the two levels, at the line of de- 
parture, a short ramplike piece of ground 

13 165th RCT Jnl, 30 Jun 44, Msgs 3, 6, 13, 14; 
165th RCT Rpt of Action Saipan, p. 5. 

14 165th RCT Jnl, 20 Jun 44, Msg 14. Also, see 
attached overlay in 27th Inf Div G-3 Jnl. 

16 105th RCT Jnl, 20 Jun 44, Msg 12. 

served as an approach to the higher plain 
from the west, but as one proceeded 
farther southward the ramp became pro- 
gressively steeper and finally developed into 
sheer cliff. Originally, A Company was de- 
ployed across this ramp from top to bottom 
with Company C tied in on the flat land 
to its right. As the advance progressed it 
would be necessary for Company A to keep 
edging more and more to the right until 
eventually it would end up on the level 
ground at the foot of the ridge. This neces- 
sarily would make effective contact with 
Company I on the left impossible. Thus it 
was that the action of the ist Battalion, 
105th, was to all intents and purposes in- 
dependent of that of the 165th on its left. 

Immediately in front of K and F Com- 
panies, 165th Infantry, there was nothing 
but open cane field sloping gently down to 
the bay on the left. Ahead of Company G, 
however, was a rubble of coral topped with 
the thick undergrowth that lined Magi- 
cienne Bay, Approximately 800 yards 
ahead of the line of departure the ground 
in front of the 165th sloped upward to a 
hill. On the left of the 3d Battalion zone 
the incline was gradual, but on the right 
of the ad Battalion the slope gave way to 
an abrupt cliff — the face of Mount Nafu- 
tan itself. 16 

Although the original jump-off hour had 
been set at 1000, General Smith found it 
necessary to postpone it to 1 1 1 5 and later 
to 1200 in order to permit the ist Bat- 
talion, 165th, to relieve the other two bat- 
talions, which were still in position along 
Magicienne Bay north of the line of de- 
parture. 17 At 1 145 division artillery laid 
down a concentrated fire along the whole 
front, particularly along the hill that 

111 Love, Battle for Saipan, pp. 1 14-15. 

17 27th Inf Div G-3 Jnl, 20 Jun 44, Msgs 9, 17. 



crossed the i65th , s line of advance. Then 
Company C, 88th Chemical Battalion, 
which had been brought up to lend general 
support to the attack, fired its 4.2-inch 
chemical mortars and set up a smoke 
screen. Six tanks from the 766th Tank 
Battalion supported the 3d Battalion, 
165th, in the center of the regimental line. 
Promptly at 1200 the troops jumped off. ls 

On the right of the regimental line the 
1 st Battalion, 105th Infantry, almost im- 
mediately came under machine gun fire 
from its left front and flank, while the right 
flank received some fire from a heavy flat- 
trajectory weapon. The whole line stopped, 
and C Company on the right dug in. 
Colonel O'Brien, the battalion commander, 
came up in an effort to locate the source 
of enemy fire and finally determined that 
it came from a small group of buildings 
almost on the battalion boundary line. 
Company A immediately put automatic 
weapons fire into this whole area. This 
seemed to stop the fire, and Colonel 
O'Brien went out to make a reconnais- 
sance. He had moved over into the build- 
ings when snipers began opening up on 
him from various houses. O'Brien immedi- 
ately ordered all the buildings burned 
down. For the next hour the battalion was 
held up while the settlement was burned 
to the ground, tanks, self-propelled mounts, 
antitank guns, and flame throwers joining 
in the arson. 

Upon completion of this task, the 1st 
Battalion, 105th, pushed forward again 
and for the rest of the afternoon ran into 
no trouble except occasional small arms 
fire. Contact with Company I on the left 

was lost during the burning of the settle- 
ment and was not regained for the rest of 
the afternoon, chiefly because of the gradu- 
ally rising ridge that now separated the 
two battalions. When the 1st Battalion dug 
in at nightfall, it had advanced about 500 
yards. 1 " 

In the zone of the 3d Battalion, 165th 
Infantry, Company I, led by three tanks, 
fell under enemy fire almost immediately 
and remained stationary for about an hour 
and a half. Its tanks proved to be more of 
a hindrance than a help since they drew 
enemy artillery fire into the area of advance 
but could not be controlled by the infantry 
because of radio failure. On the left, Com- 
panies K and F were faring considerably 
better, and at 1405 Company K reported 
that it was 400 yards ahead of I Company 
and out of contact. 20 Meanwhile, on the ex- 
treme left of the regimental line G Com- 
pany was stopped by a nest of Japanese 
hidden in the underbrush near the ocean 
shore and made no further advance. With 
both flanks of the line retarded, the two 
battalion commanders ordered their re- 
serve Companies, E and L, to take po- 
sitions on the extreme left and right, 
respectively. These moves were completed 
about 1630, and the regiment prepared to 
continue the advance. 21 

Heavy mortar fire was laid down, and 
both battalions jumped off in a continu- 
ance of the attack. On the regimental 
right progress was slow since the entire 3d 
Battalion had to contend with the heavy 
undergrowth and was moving up hill. On 
the left, E Company commenced to receive 
considerable fire from the hills north of 

18 165th RCT Jnl, ao Jun 44, Msgs 45, 47, 48, 
50, 51 ; 105th RCT Jnl, ao Jun 44, Msg 23; Roy 
E. Appleman, Army Tanks in the Battle for Sai- 
pan, p. 18, MS in OCMH. 

19 Love, Battle for Saipan, pp. 134-26. 
29 3d Bn 165th RCT Jnl, 20 Jun 44, Msg 38. 
21 ad Bn 165th RCT Jnl, 20 Jun 44, Msg 28; 3d 
Bn 165th RCT Jnl, 20 Jun 44, Msg 49. 



Mount Nafutan and was pinned down. By 
1730 no further progress seemed possible 
before nightfall, and all units were ordered 
to dig in for the night. Company E with- 
drew about a hundred yards before doing 
so. 22 Casualties had been relatively light, 
the 105th suffering only one man killed 
and five wounded; the 165th, six killed, 
twenty-one wounded, and one missing in 
action/ 3 

Meanwhile, the 3d Battalion, 105th In- 
fantry, which was still under control of its 
parent regiment, had been pushing east- 
ward along the southern shore. Little 
infantry opposition was encountered by 
either of the two assault companies, al- 
though they did receive scattered artillery 
fire at different times during the day. By 
nightfall the battalion had reached a point 
only a hundred yards short of tying in with 
the attack coming down Nafutan peninsula 
from the north. The division line, there- 
fore, presented an almost solid front that 
hemmed the southern defenders of the 
island into an ever-tightening pocket. 21 

During 20 June the 106th Infantry Reg- 
iment landed on Saipan and was assigned 
as corps reserve. 25 As soon as the regiment 
was ashore the 2d Battalion, 105th Infan- 
try, was released to the control of the 27 th 
Division, and General Ralph Smith im- 
mediately ordered it to assemble in division 
reserve at the southwest corner of Aslito 
airfield. 2 * 5 

Along the division's front line that night 
there was little activity except in the center 
in the zone of the 3d Battalion, 165th In- 
fantry. Shortly before 2200 enemy guns 

22 Love, Battle for Saipan, pp. 122-24. 

23 27th Inf Div G-i Periodic Rpt 4. 

24 Love, Battle for Saipan, p. 126. 

25 1 06th RCT Jnl, 20 Jun 44, Msg 1 1. 

2(i 27th Inf Div G-3 Jnl, 20 Jun 44, Msg 42. 

began opening up not more than 1 50 yards 
to the direct front. The fire was point 
blank and was aimed at both the 3d Bat- 
talion zone and the area held by the 1st 
Battalion, 105th Infantry. In the zone of 
the former, the Japanese guns overshot 
their mark and no damage was done, but 
during the barrage some rounds fell on A 
Company, 105th, killing one man and 
wounding three. 27 

Action of 21 June 

Plans for 21 June called for a continu- 
ance of the attack to the south on Nafutan 
Ridge. At a conference held at the 27th 
Division command post at 2200 on 20 
June, the plan was reaffirmed, but with 
some changes. At Colonel Kelley's request, 
General Ralph Smith ordered the fresh 2d 
Battalion, 105th Infantry, to relieve the 2d 
Battalion, 165th, on the left of the line as 
early as possible the next morning. The at- 
tack was to jump off at 0930 after a 
thirty-minute artillery preparation. Upon 
reaching the first phase line, where the 
3d Battalion, 165th Infantry, was to be 
pinched out, control of the attack south- 
ward was to be assumed by Colonel Bishop 
of the 1 05th Infantry. Field Order Number 
45, 27th Infantry Division, which con- 
tained these plans, was issued at 0615, 21 
June/" By 0900 the 2d Battalion, 105th, 
had relieved the 2d Battalion, 165th, on 
the left of the line. 29 As the action opened, 
then, on the morning of 21 June, the 27th 
Division units on the line from right to 
left (west to east) were: Companies L, 
I, C, and A, 105th Infantry; Companies 

27 Love, Battle for Saipan, p. 128. 
2n 27th Inf Div FO 45, 21 Jun 44. 
29 165th RCT Jnl, 21 Jun 44, Msg 21. 



Tnkantryman Milking an Island Goat in Nafutn.n area, ao June. 

L, and K, 165th; and Companies G and 
F, 105 th. 

On the extreme right, the 3d Battalion 
of the 1 05th, still pushing its way eastward 
along the southern coast of Saipan, met 
serious enemy opposition for the first time. 
Shortly before noon the right platoon of 
Company I, operating along the seashore, 
crossed the face of a cave in the ridge and 
a Japanese machine gun opened up, plac- 
ing enfilade fire all along the platoon line. 
The advance stopped at once. On request of 
the company commander, division dis- 
patched a platoon of tanks from those that 
had come ashore with the 106th infantry. 
In the meantime, Lt. Col. Edward T. Bradt, 
battalion commander, sent forward a self- 
propelled mount from the Cannon Com- 
pany. The vehicle sprayed the area with 

fire but failed to get close enough to the 
cave to deliver direct fire into its mouth. 
Shortly after 1500 the tanks arrived and 
immediately knocked out the position with 
their machine guns and 37-mm's. The bat- 
talion line then remained stationary while 
a loud speaker was sent forward from di- 
vision headquarters in an unsuccessful at- 
tempt to persuade the remaining Japanese 
troops and/or civilians to come out of the 
cave. Shortly before 1700 the battalion line 
pushed forward to a point about 600 yards 
from the morning line of departure and 
dug in for the night. Contact with the 1st 
Battalion, 105th, on the left had been lost 
during the day's movement, and a 
small gap remained between the two bat- 
talions. 30 

30 Love, Battle for Saipan, p. 154. 



In the zone of the ist Battalion, 105th, 
Company A on the battalion left lost 
ground even before the drive got under 
way. At daylight, when it became evident 
from the sound of firing on its left that the 
enemy had entrenched himself firmly to 
the front, Capt. Louis F. Ackerman or- 
dered his men to move back about a 
hundred yards to a less exposed position. 
A further backward movement to adjust 
its lines to those of the unit on its left 
brought Company A a full 200 yards be- 
hind the positions where it had dug in the 
night before. 

After jumping off at 0930, the ist Bat- 
talion, 105th, moved forward slowly with- 
out opposition. The advance was delayed 
to permit cane fields to be burned to the 
company's front, and even two hours after 
the jump-off Company A had not yet 
reached the place where it had bivouacked 
the night before. Finally, at 1255, wnen 
Company A had advanced about a hun- 
dred yards ahead of the line where it had 
spent the previous night, it was hit by a 
heavy mortar concentration coupled with 
sweeping small arms and automatic weap- 
ons fire. This caught the advancing troops 
in open terrain without cover, Ackerman 
immediately radioed battalion headquar- 
ters for tanks and ordered his men back 
into the foxholes of the night before. 

On the right of the battalion line, Com- 
pany C had guided its advance on Acker- 
man's company. Most of the men in this 
part of the line had better protection than 
did A Company, so when the mortar bar- 
rage hit, ist Lt. Bernard A. Tougow, in 
command of C, kept his men on the line. 
Within a few minutes after A pulled back, 
Colonel O'Brien, battalion commander, 
arrived at the C Company command post 
with three tanks, which immediately went 

to work to break up a small Japanese 
counterattack. The tanks then moved over 
to the left to meet Captain Ackerman, who 
had put in the request for their assistance. 

O'Brien organized a co-ordinated attack 
along the whole front of his battalion and 
supported it by the tanks, which he placed 
in front of Company A. Shortly before 
1500 the assault moved off after a brief 
artillery preparation. The tanks, which 
were buttoned up, moved out ahead of the 
line of infantrymen for a few minutes, then 
veered to the left and finally reversed their 
course and headed back toward the Amer- 
ican line firing as they came. Colonel 
O'Brien's frantic efforts to contact the 
tankers by radio failed, and he finally ran 
out in the midst of this fire to meet them. 
Crawling up on the turret of the first tank 
he met, he banged on it with his pistol 
butt. The tank then contacted the other 
two by radio and the firing stopped 
momentarily. O'Brien turned the vehicles 
around and then took up a position atop 
the lead tank's turret and ordered the ad- 
vance to proceed. 

The whole battalion jumped off in a 
rapid push that carried it across the open 
ground. Throughout the movement most 
of the men advanced at a dogtrot behind 
the tanks, keeping up a steady fire to the 
front. O'Brien continued to ride the tank 
turret of the lead tank, giving directions 
to the men inside with his pistol butt and 
waving the infantrymen forward. During 
the advance A Company lost two men 
killed and three wounded. Company C on 
the right suffered no casualties. 31 

In the center of the division line, Com- 
panies L and K of the 165th Infantry 
jumped off on schedule at 0930. They had 

111 Ibid., pp. 147-50; 105th RCT Jnl, ai Jun 44, 

Msgs 19, 21, 27. 



made some progress by 1255, when they 
were held up by a heavy concentration of 
mortar fire, most of which landed in the L 
Company area. Within the space of a few 
minutes one man was killed and eleven 
were seriously wounded; then the barrage 
ceased as abruptly as it had begun. By that 
time all of the 3d Battalion was badly dis- 
organized and made no further advance 
during the afternoon. This left L Company 
of the 165th some 500 yards to the left rear 
of Company A, which had advanced rapid- 
ly during the afternoon with the aid of the 
tanks under Colonel O'Brien's personal 

To fill the gap, O'Brien ordered in the 
1st Platoon of his reserve Company B. The 
platoon leader sent out a patrol that re- 
ported that a number of Japanese had 
taken up position with a machine gun at 
the crest of the ridge between the two bat- 
talions and that the only way firm contact 
could be established was by knocking out 
the position. O'Brien then ordered the pla- 
toon to face the ridge, deploy, and assault 
it frontally from the west. After a short 
mortar concentration the platoon attacked 
at 16 1 5, but was immediately pinned down 
by enemy fire that killed two men and 
wounded three others. Shortly afterward, 
O'Brien received an urgent radio message 
indicating that Company L was being 
fired on from the direction of the 1st Pla- 
toon, Company B. The assault on the west 
face of the ridge was promptly called off 
and the gap along the battalion line re- 
mained unclosed for the night. 32 

The most serious difficulties of the day's 
fighting for Nafutan came on the extreme 
left of the division line. Here, the un- 
blooded 2d Battalion, 105th Infantry, was 

:r2 Love, Battle for Saipan, pp. 145-46, 151-52. 

inserted in the line with Company G on 
its right and Company F on its left close to 
the ocean shore, while the reserve company 
took position north of the front line along 
the eastern coast. 

The terrain to the immediate front was 
extremely difficult. The most prominent 
feature was the nose of Mount Nafutan, a 
sheer cliff splitting the battalion front like 
the bow of a ship. The cliff was not more 
than thirty feet high, but the approach to 
it was up a steep slope through the stubble 
of a cane field that offered no cover. 

The battalion jumped off on schedule at 
0930. On the right, Company G was 
immediately hit by enemy small arms, ma- 
chine gun, and mortar fire. One source of 
the fire was soon found to be a machine 
gun located on the top of the cliff. Capt. 
Frank Olander, company commander, or- 
dered one squad to assault the cliff itself, 
but the men had no sooner reached the 
top of the cliff than they were recalled be- 
cause of their precarious and isolated 
situation. A second squad was sent to the 
top, but the underbrush was so thick that 
the men failed to spot the critical machine 
gun. Another platoon that had meanwhile 
attempted to infiltrate around to the right 
of the nose of Mount Nafutan was soon 
pinned down by the enemy fire from the 
top. The squad on top of the cliff was then 
called back, and the company commander 
made his way to the battalion command 
post to request more aid. 

Lt. Col. Leslie M. Jensen, the battalion 
commander, immediately ordered two self- 
propelled mounts from the 165th Infantry 
Cannon Company (the 105th Cannon 
Company was not yet ashore) to carry ra- 
tions and water to the isolated men around 
the base of the cliff. He then called division 
artillery for help but was advised that a 



concentration on the nose of the ridge or 
near any part of Jensen's front line was 
inadvisable because of the advanced posi- 
tions of the rest of the American line. 
Eventually, however, it was agreed that 
one battery of 105-mm. howitzers could be 
sent forward to fire point-blank against the 
cliff. Olander was instructed to withdraw 
his men under cover of this fire. The de- 
cision was reached at 1535, and Battery 
B, 104th Field Artillery Battalion, was or- 
dered to the front lines at once. 33 

For the next hour, under Olander's in- 
structions, Company G attempted no fur- 
ther movement. Then, under cover of the 
artillery barrage the entire company drew 
back to positions to the rear of that morn- 
ing's line of departure. 34 

On the extreme left, close to the ocean 
shore, Capt. Earl White, commanding 
Company F, 105th Infantry, sent his 2d 
Platoon south through the scrub fringing 
the shore line with instructions to search 
for a route that would lead them onto 
Mount Nafutan from the rear. At 1700, 
after an afternoon of extremely difficult 
work through the coral and underbrush, 
the platoon finally gained the top of Mount 
Nafutan. During this period, White had 
kept mortar fire on the Japanese positions 
on top of the ridge that were holding back 
the rest of his company. Sometime during 
the afternoon the Japanese appear to have 
picked up their gum and moved out. When 
the 2d Platoon arrived on top of the ridge 
at 1700 the men found it unoccupied, but 
within a few minutes of the arrival of the 
platoon the company commander ordered 
it to return to its starting point. Upon the 
return of the platoon, White ordered Gom- 

s;J 165th RCT Jnl, 21 Jun 44, Msg 121. 
: ' 4 Love, Battle for Saipan, pp. 137-45. 

pany F to pull back behind the line of de- 
parture where there was better cover and 
where Company G had already dug in. 35 
Thus, by the close of the fighting on 21 
June, troops of the 27th Division had made 
insignificant progress on either flank of its 
attack down Nafutan Point, but had 
made a slight dent in the center. 3 *"' The in- 
termediate objective line about halfway 
down the peninsula from the original line 
of departure of 20 June was still from five 
hundred to a thousand yards away. The 
nose of Mount Nafutan, which had been 
reached by elements of Companies F and 
G, 105th Infantry, had been relinquished 
and the mountain itself still had to be 
scaled before the southeastern tip of the 
island could be secured. Casualties for the 
day's fighting on Nafutan came to seven 
killed and fifty-seven wounded in action. 37 

Change of Plan: Relief of the 
165th Infantry on Nafutan Point 

By 21 June the two Marine divisions 
had completed their pivoting movement to 
the north, and General Holland Smith pre- 
pared to launch a full-scale attack against 
the Japanese main line of resistance in that 
area. To do so, he would need the bulk of 
the 27th Division as corps reserve and, ac- 
cordingly, he decided to reduce the number 
of troops committed to Nafutan Point and 
to remove most of the men to the reserve 
area behind the Marine front lines to the 
north. His opinion that these troops would 
no longer be needed on Nafutan was re- 
inforced by a report from 27th Division 

3i Ibid., pp. 136-37. 

30 See attached Hq 27th Inf Div Situation 
Overlay to Accompany G-3 Periodic Rpt 4, Sai- 
pan, 1/20,000, 21 1600 Jun 44. 

37 27th Inf Div G-r Periodic Rpt 5. 



headquarters stating that the only enemy 
left in that area consisted of 300 to 500 
service personnel from the remnants of 
naval air units originally stationed on As- 
lito, plus a larger number of civilians. 38 

Hence, on 2 1 June Holland Smith issued 
his operations Order Number 9-44, which 
was received at 27 th Division Headquar- 
ters at 1215 that day. 38 The 27th Infantry 
Division (less one infantry battalion and 
one light tank platoon) was to assemble 
northwest of Aslito airfield in corps reserve. 
Division artillery was to pass to control of 
the XXIV Corps Artillery. One infantry 
battalion (undesignated) of the division 
was ordered to remain in the garrison area, 
that is, Nafutan peninsula. "It will mop 
up remaining enemy detachments, main- 
tain anti-sniper patrols . . . and protect 
installations within its zone of action with 
particular attention to ASLITO Airfield." 

The slow progress that his division had 
made on the afternoon of the 21st, how- 
ever, convinced General Ralph Smith that 
more than a single battalion would be 
necessary to clean up the point. According- 
ly, at 1435, his headquarters notified Col. 
Robert Hogaboom, USMC, G-3 of North- 
ern Troops and Landing Force, that at 
least two battalions would be needed for 
the next day's operations in that area. 40 
At 1700 General Ralph Smith called Gen- 
eral Holland Smith and recommended that 
all of the 105th Regimental Combat Team 
be left in the Nafutan Point area. General 
Holland Smith agreed to this but stipulated 
that only two of the io5th's battalions be 
used there. The other would be held in 

reserve ready for use elsewhere if neces- 
sary, 41 

This modification of Operations Order 
Number 9-44 was contained in a mail 
brief issued by General Holland Smith that 
arrived at 27th Infantry Division headquar- 
ters at 0830 on 22 June. In the words of 
the message, "1 RCT will continue mission 
in Garrison Area [Nafutan] of cleaning 
up remaining resistance & patrolling 
area." 42 The order did not designate specif- 
ically which regimental combat team was 
intended, although the previous day's con- 
versation had clearly indicated that the 
105th was to be used for the mission. 

At 2000, 21 June, after his conversation 
with General Holland Smith but before re- 
ceiving the mail brief modifying the latter's 
original orders, General Ralph Smith is- 
sued his Field Order Number 45-A, which 
contained the following instruction to the 
105th Infantry: 

RCT 105 will hold present front line facing 
NAFUTAN PT, with two Battalions on the 
line and one Battalion in Regimental Re- 
serve. It will relieve elements of RCT 165 
now on the present front line by 0630 22 
June. The Battalion in reserve will not be 
committed to action without authority from 
the Division Commander. Reorganization of 
the present front line to be effected not later 
than 1 100 33 June and offensive operations 
against the enemy continued. Reserve Bat- 
talion will maintain anti-sniper patrols in the 
vicinity of ASLITO AIRFIELD. 43 

The wording of this paragraph and the 
fact that it was issued at all to the 105th 
Infantry by 27th Division's commanding 
general was soon to become a major bone 
of contention between Generals Holland 

ss 27th Inf Div Gva Periodic Rpt 5, 21 Jun 44. 
3M NTLF Opn Order 9-44, 21 Jun 44; 27th 
Inf Div G-3 Jnl, 2 1 Jun 44, Msg 34. 

i0 27th Inf Div G-3 Jnl, 21 Jun 44, Msg 43. 

41 Ibid., Entry 65. 

12 27th Inf Div G-3 Jnl, 22 Jun 44, Msg 14. 
43 27th Inf Div G-3 Rpt, Battle for Saipan, FO 
45, Able, aooo, 21 Jun 44. Author's italics. 



Smith and Ralph Smith and was one of the 
alleged reasons for the latter's being sub- 
sequently relieved of his command. 

Action of 22 June 

22 June was spent reorganizing the front 
lines facing Nafutan Point. 44 On the right 
General Ralph Smith ordered the 3d Bat- 
talion, 105th Infantry, to hold its line and 
spread out to the left to relieve the 1st 
Battalion of the same regiment, which was 
to revert to corps reserve. On the left, the 
2d Battalion, 105th, was to hold its line 
facing Mount Nafutan and move to the 
right to relieve the 3d Battalion, 165th In- 
fantry. On the right, the 3d Battalion ac- 
complished its assigned relief mission, but 
on the left the 2d Battalion ran into 

During the preceding night it had be- 
come evident that the Japanese were 
preparing positions on the nose of Mount 
Nafutan, and Captain Olander, G Com- 
pany commander, requested permission to 
pull his men back to less exposed positions. 
Permission was granted, but before the 
move could be executed the enemy on 
Mount Nafutan opened fire with machine 
guns, small arms, and mortars, killing seven 
men and wounding twenty-one. 

Companies G and F immediately pulled 
back a considerable distance to the rear 
for reorganization, leaving E Company to 

44 The following narrative of the Nafutan Point 
action from 22 to 28 June is derived mostly from 
information gained by Capt. Edmund G. Love and 
set forth in his manuscript, Battle for Saipan. 
During most of this period only one battalion 
was engaged at Nafutan, the 2d Battalion, 105th 
Infantry. All of the records of the battalion were 
destroyed by fire on 7 July, and the records of 
higher echelons contain very little detailed in- 
formation concerning activity in the area. See 
Love, Battle for Saipan, p. i6arT. 

prevent any breakthrough. Company G, 
which had been badly hit on the a ist as well 
as on the morning of the 22d, took more 
than two hours to reorganize. By 0946 Cap- 
tain Olander was ready to move again, but 
by this time his company had four officers 
and only seventy-two enlisted men, less than 
half of its original strength. With these few 
soldiers he was expected to take over a zone 
then held by a full battalion. The reorgani- 
zation had taken place some 400 yards 
behind the position of the night before and 
the men now marched another 600 yards 
to the original line of departure from which 
the attack had jumped off on 20 June. 
From this point the company commander 
moved his men up to relieve the 3d Bat- 
talion, 165th, at 1025. 45 Because his com- 
pany was understrength, Olander had to 
place his men some twenty yards apart in 
the skirmish line. 

The 2d Battalion, 105th, spent the rest 
of the afternoon reorganizing its line, and 
made no further advance. By nightfall, in 
fact, there was a net loss of ground on the 
2d Battalion front. 

Meanwhile, about 15 15 General Ralph 
Smith visited headquarters of Northern 
Troops and Landing Force to consult with 
the corps commander about plans for the 
immediate future. General Holland Smith 
expressed his concern regarding the slow- 
ness of the advance on Nafutan Point. He 
said that "he did not wish to be unreason- 
able but that Colonel Bishop [Command- 
ing Officer, 105th Infantry] must not be 
permitted to delay. If he couldn't do it, to 
send somebody who could. 5 ' 46 In response, 
General Ralph Smith "pointed out difficult 
terrain and Jap positions in caves and said 

45 3d Bn 165th RCT Jnl, 22 Jun 44, Msg 14. 
4,i Ralph Smith, Notes, Saipan, p. 9. 



rapid advance was impracticable if undue 
losses were to be avoided and if Japs were 
to be really cleaned out. [He] said that 
continuing pressure would be applied and 
that [he] thought the point could be 
cleaned in a couple of days more." 47 

Shortly after this meeting, General 
Ralph Smith went to see General Erskine, 
Holland Smith's chief of staff. General Ers- 
kine apprised him of the corps plan to 
pass the 27th Division between the two 
Marine divisions on the northern front. As 
to Nafutan Point, Erskine expressed his be- 
lief that one battalion could finish up the 
job there. 48 

As a result of these afternoon confer- 
ences, General Ralph Smith returned to 
the division command post and drew up 
Field Order Number 46, which was issued 
at 2100. In part, the order read: "ad Bat- 
talion, 105th Infantry (1 Platoon Light 
Tanks attached) [will] continue opera- 
tions to mop-up remaining enemy detach- 
ments in NAFUTAN POINT area. On 
completion of this mission, [it will] revert 
to Corps control as Corps Reserve." 40 

An hour later Holland Smith issued 
Operations Order Number 10-44, which 
was received at 27th Division command 
post at 233o. S0 In reference to Nafutan 
Point this order read: "2d Battalion, 105th 
Infantry (with one light tank platoon at- 
tached) [will] continue operations at 
daylight to mop up remaining enemy 
detachments in NAFUTAN POINT area. 
Upon completion this mission [it will] re- 
vert to Corps control as Corps reserve." 

Although there was no serious discrep- 

47 ibid. 

ifi Ibid., p. 11. 

49 27th Inf Div FO 46, 2100, 22 Jun 44, in 27th 
Inf Div G-3 Rpt. 

50 NTLF Opn Order 10-44, 2200, 22 Jun 44; 
27th Inf Div G-3 Jnl, 2 2 Jun 44, Entry 78, 

ancy between these two sets of orders, 
General Holland Smith was becoming 
alarmed over the fact that the battalion 
on Nafutan Point was getting orders from 
two different sources. Next day he warned 
General Ralph Smith: "2d Battalion, 105th 
by my operations order 10-44 not un( ier 
your tactical control and should not be in- 
cluded in your tactical orders. Please take 
steps to rectify." ni 

Later in the operation, in requesting the 
relief of General Ralph Smith, General 
Holland Smith alleged that Field Order 46 
"contravened the NT and LF order by is- 
suing tactical orders to the 2d Battalion, 
105th Infantry, to continue operations to 
mop up enemy resistance in NAFUTAN 
POINT area. The 2d Battalion, 105th In- 
fantry, by NT and LF Order No. 10-44, 
had been removed from the tactical control 
of the 27th Infantry Division." n2 

Stalemate on Nafutan 
2^24 June 

As a result of these new orders the job 
of finishing off Nafutan peninsula was left 
to a single rifle battalion supported by one 
platoon of six light tanks. The battalion 
was to have no assistance from artillery ex- 
cept for whatever fire support might be 
provided by naval ships operating in the 
area. The front line currently held by the 
American troops ran along the northern 
base of the peninsula for a distance of 
roughly 2,500 yards. The terrain was 
mountainous, full of cliffs, crevices, and 
caves. Yet, it must be added that, because 
of the shape of the peninsula, any contin- 
uous forward advance of the attacking 

51 NTLF G-3 Jnl, 23 Jun 44, M S g 23141 1. 
53 CG Expeditionary Troops to CO TF 51, 24 
Jun 44, Buckner Board Rpt, p. 2 and Exhibit D. 



troops would automatically reduce the 
length of the front and thereby shorten the 
line. In effect, the troops were moving 
down an inverted isosceles triangle from 
base to apex. An advance of a thousand 
yards along the axis of the attack would 
reduce the front from approximately 2,500 
yards to approximately 1 ,000 yards. 

Nevertheless, General Ralph Smith was 
sufficiently alarmed at the wide dispersion 
of the troops left along the front line on 
Nafutan to warn General Holland Smith 
of the possible consequences. "I want to 
draw your attention," he wrote on 23 
June, "that it is within the enemy's cap- 
abilities at NAFUTAN Point to infiltrate 
small bodies of men through our lines at 
night and execute considerable damage to 
the planes and supplies at Conroy [that is, 
Aslito or Isely] field." He added that the 
Seabees and Air Forces troops working on 
the field should be alerted and would have 
to provide their own local security against 
enemy groups that might infiltrate through 
the lines of the 2d Battalion, 105th In- 
fantry. 53 

General Holland Smith had withdrawn 
the bulk of the troops previously com- 
mitted to Nafutan because they were more 
urgently needed in the north and because 
his intelligence reports indicated that the 
number of Japanese remaining on the point 
was small and probably ill-equipped. Two 
days earlier the intelligence sections of both 
the 27th Division and the 105th Infantry 
had estimated that only from three to five 
hundred enemy service personnel remained 
bottled up in that area, and no revision of 
that estimate had been made since. 54 Actu- 

ally, as later events were to prove, the 
number was much larger, but as of the 23d 
no responsible authority had issued any re- 
port to indicate that this was so. 

The change in orders now necessitated 
another shuffling of the line. Lt. Col. Leslie 
Jensen, commanding officer of the 2d Bat- 
talion, 105th, ordered E Company to 
march to the right and relieve the 3d Bat- 
talion, 105th. Company G was shifted to 
the right so that part of the company was 
below the 300-foot ridge line west of 
Mount Nafutan and part was on the north 
slope of the ridge. This reorganization was 
completed at 1230, and the company was 
ordered to attack at 1400. Company F, 
which on the morning of 23 June was still 
in position before the nose of Mount Nafu- 
tan, was ready to attack at 0800, but, be- 
fore the attack could be launched, was 
withdrawn and reorganized. Colonel Jen- 
sen then waited until he saw how far his 
other two companies would stretch before 
ordering F back into the line. Thus, in spite 
of Holland Smith's orders to "continue 
operations at daylight," the 2d Battalion 
spent the entire morning trying to readjust 
its lines to stretch clear across Nafutan 
Point. When this readjustment was com- 
pleted, the three companies were in position 
in a broken line with Company E on the 
right, G in the center, and F on the left. 55 

On the right (west) flank, one platoon 
of Company E managed to push through 
the coral fringing the beach for a distance 
of about 300 yards without any opposition. 
However, at the day's end this advance 
platoon was pulled back to its starting 
position because Capt. Clinton F. Smith, 
the company commander, had not been 

53 Ralph Smith to CG NTLF, 23 Jun 44, Buck- 
ner Board Rpt, Exhibit VVV. 

84 105th Inf FO 30, jfloo, 21 Jun 44; 27th Inf 
Div G-2 Periodic Rpts 5, 6, 7. 

55 Unless otherwise noted, the account of the 
actions of the 2d Battalion, 105th Infantry, is de- 
rived from Love, Battle for Saipan, pages 1 75-98, 



able to establish contact with G Company 
on his left and was fearful of infiltration. 

Meanwhile Company G moved up to the 
top of Ridge 300. There it came under fire 
from at least four machine gun positions to 
its left (east). Captain Olander ordered 
his men not to return fire for fear of en- 
dangering the men of F Company, who 
were presumably operating in the general 
area from which the enemy was firing. 
Efforts to bring up the three light tanks 
attached to the unit failed because of the 
precipitous coral terrain, and finally the 
company commander ordered his men to 
withdraw to the bivouac area of the night 

In the zone of Company F, the qd Pla- 
toon reached the top of Mount Nafutan by 
skirting it to the left through the brush 
just inland of the east coast and coming 
up to it from the rear. The men met no 
opposition en route. The 1st Platoon was 
ordered to move up the valley between 
Ridge 300 and Mount Nafutan. For about 
an hour it proceeded without any opposi- 
tion, but suddenly the whole column came 
under fire from a machine gun on the right 
in the direction of Ridge 300, Three tanks 
were called up and for better than half an 
hour these vehicles sprayed the hills on 
both sides of the valley. Nevertheless, at 
1700 Captain White, the company com- 
mander, called the platoon back out of the 
valley and ordered it to dig in along the 
morning's line of departure. Meanwhile, 
the 3d Platoon had moved along the inside, 
east of Ridge 300, with no opposition until 
about 1500. There it halted and waited for 
the rest of the company to move abreast. 
When this failed to happen, it too with- 
drew to dig in for the night with the rest 
of F Company. 

Thus at the close of the day the 2d Bat- 
talion, except for one platoon atop Mount 
Nafutan, had withdrawn to approximately 
the same positions it had occupied at the 
beginning of the day's advance. The bat- 
talion was dug in in four widely separated 
perimeters with no contact between them. 
The perimeter of E Company on the right 
was about 1,000 yards from that of G in 
the center; G, in turn, was about 800 yards 
from the Company F positions, while one 
platoon of F was in an inaccessible position 
another 800 yards to the left front. 

As before, General Holland Smith's or- 
ders for 24 June called for the 2d Battalion, 
105th Infantry, to "continue operations at 
daylight to mop up remaining enemy de- 
tachments in NAFUTAN POINT area." 56 
However, not until 0800 did the battalion 
actually commence moving. The interim 
was spent trying to readjust the overex- 
tended lines of the battalion and trying to 
establish at least a semblance of contact 
between the various units. 

On the extreme right flank, two pla- 
toons of Company E, against almost no 
opposition, reflushed the area they had 
traversed the day before and by nightfall 
reached a point about 100 yards beyond 
that previously gained. The 3d Platoon of 
this company, on the left, ran into more 
difficulty. During the early part of the 
morning the unit had moved to the left 
and re-established contact with G Com- 
pany. By 1000, without running into any 
Japanese, it had reached the point of its 
furthest advance of the day before. Shortly 
thereafter the leading squad was hit by 
heavy small arms and machine gun fire 
from its left rear and was forced to take 
cover in a group of small houses. Mean- 

36 NTLF Opn Order 1 1-44, 23 Jun 44. 



while, the other two squads to the rear 
laid mortars on the suspected source of 
enemy fire, but failed to knock it out. Then, 
about 1500, a force of from fifty to 
seventy-five Japanese rose up out of the 
ground and launched a counterattack 
through the gap that had developed be- 
tween E and G Companies. With this, all 
further progress ceased, and shortly after 
dark the entire platoon moved back to the 
company perimeter of the night before. 

G Company in the center was late in 
moving out. Captain Olander waited until 
Company E on his right had made contact 
and until F on his left had been reorgan- 
ized. He then further delayed his jump-off 
until the arrival of the three light tanks he 
had requested. Moving off about n 30 
Company G quickly recovered the ground 
it had taken the previous day, and then it 
again ran into machine gun fire. A tank 
was brought forward, succeeded in locating 
one of the enemy guns, and in a few min- 
utes silenced it with 37-mm. fire. Shortly 
after this the Japanese counterattack on 
the right developed, and although G Com- 
pany was not hit, it remained stationary 
for two hours. 

At approximately 1630 Olander once 
again ordered his company to advance. 
Four enemy machine gun positions in the 
immediate front were taken out by tank 
guns. The reduction of these positions put 
the company ahead of the units on the 
right and left, and Captain Olander swung 
his men to the left in an attempt to take 
out a group of machine guns that were 
holding back the advance of Company F. 
This move was effected in spite of ap- 
proaching darkness, and within a few 
minutes after making the turn G Company 
surprised a pocket of about fifty Japanese 
and wiped them out within ten minutes. 

In the ensuing darkness, however, all 
organization within the company broke 
down. Olander lost contact with his pla- 
toon leaders, and the latter pulled their 
men back to the bivouac area of the night 

The action of Company F on the left was 
in general a repetition of that of the pre- 
vious day. The 2d Platoon, which had 
spent the night on Mount Nafutan, was 
ordered to build up a skirmish line and 
comb the nose of the ridge until the 1st 
Platoon could move up on its right. How- 
ever the latter unit, while en route to the 
top of Mount Nafutan, ran into scattered 
rifle fire and stopped in its tracks. Mean- 
while, on the company's right, the 3d Pla- 
toon was held up by a Japanese machine 
gun. A self-propelled mount from the 
105th Cannon Company knocked this po- 
sition out, but retired before disposing 
of a second machine gun, which had 
wounded one of its crew. The platoon 
leader then sent out a squad to get the 
weapon, but a third gun opened up and 
pinned the squad down. By this time night 
was approaching and, as no further prog- 
ress seemed likely, Captain White ordered 
his entire company including the platoon 
on top of Mount Nafutan to withdraw to 
the G Company perimeter of the night 

At nightfall then, the 2d Battalion, 
105th Infantry, occupied positions in prac- 
tically the same area in which it had dug 
in the previous night, except that the 
platoon atop Mount Nafutan had been 
recalled. At 18 18 control of the battalion 
had passed to the Army Garrison Force. 57 

57 Memo, Col Geoffrey M. O'Gonnell for Gen 
Richardson, 12 Jul 44, sub: Opns of 2d Bn, rosth 
Inf, Saipan, on Nafutan Point, Buckner Board 
Rpt, Exhibit WW. 



Since Maj. Gen. Sanderford Jarman, USA, 
the island commander, had taken com- 
mand of the 27 th Division during the 
day, command of the forces on Nafutan 
Point passed to the control of Col. Geoffrey 
M. O'Connell, General Jarman's chief of 

Nafutan Secured 
25-28 June 

Colonel O'Connell's first step to expe- 
dite the capture of Nafutan was to assign 
two batteries of 90-mm. antiaircraft guns 
and four 40-mm. guns to support the ad- 
vancing infantry. The 90-mm. guns were to 
fire from their fixed positions on Aslito 
field, and the ^o-mm.'s were ordered to 
move into forward positions in direct sup- 
port. Because of the mountainous terrain 
and the impossibility of bringing direct fire 
against most of the Japanese positions, the 
go-mm. guns were ordered to fire air bursts 
into the tree tops, approximately twelve 
feet above ground level. In the opinion of 
Colonel O'Connell, "The high muzzle ve- 
locity of these weapons, their rapid rate of 
fire and the flexibility obtainable by raising 
and lowering the height of burst made 
them particularly effective for support in 
this type of terrain." 58 The 40-mm. 's were 
used for direct fire and were to prove re- 
markably accurate in hitting cave entrances 
as small as four or five feet in diameter 
from an average range of 2,000 yards. 5 * 

O'Connell's plan for the infantry differed 
somewhat from that previously employed. 
Company E was ordered to leave its 3d 
Platoon in the area adjacent to the penin- 

sula's west coast at the point where the 
company had bivouacked the two preced- 
ing days. The other two platoons were to 
move to the left, establish firm physical 
contact with G Company, build up a 
skirmish line, and move south along the 
west slope of Ridge 300. Company G was 
to form a line on E Company's left flank 
and advance in co-ordination with that 
unit. F Company was to deploy two pla- 
toons across the mouth of the valley be- 
tween Ridge 300 and Mount Nafutan, 
while the third platoon moved along the 
east slope of the ridge in co-ordination 
with the other two companies. 60 

By 1030 of 25 June, the 1st and 2d 
Platoons of E Company had swung left 
and established contact with Company G. 
About 1 130, after an advance of nearly 
150 yards, the leading squad of the 1st 
Platoon on the right ran into a fusillade of 
fire and was pinned down. Tanks were 
called up but became entangled in the 
undergrowth and rocks and could be of no 
assistance. At 1600 the company comman- 
der ordered both platoons to retire about 
forty yards behind their farthest point of 
advance and tie in with Company G and 
spend the night. 

Company G had little or no opposition 
during the day, but its advance was slow 
because it was held up by the halting 
forward movement of the units on both 
flanks and because the tanks had extreme 
difficulty in maneuvering over the terrain. 
About noon the company reached the gun 
position it had knocked out during the late 
afternoon of the preceding day, and after a 
heavy fire by antiaircraft guns, moved on 
through it. The position contained four 

5S O'Connell, Opns in Nafutan Point Area, Sai- 
pan, 2 Jul 44, Buckncr Board Rpt, Exhibit XX, 
P- 3- 

s() Ibid.; see also, Memo cited n, 57, 

fi0 The following account, unless otherwise 
noted, is derived from Love, Battle for Saipan 
pp. 199-215. 



Skirmish Line on west slope of Ridge 300. Note man with bazooka at extreme 

heavy machine guns and two 50-mm. 
mortars. The company advanced another 
twenty-five yards but was then held up 
because of the dense growth of scrub 
brush. Captain Olander worked his tanks 
into position and' for two hours sprayed 
this area with machine gun fire and canis- 
ter. Just as he was about to continue the 
advance, the tanks notified him that it was 
1600 and they were about to withdraw. 
This notice plus the fact that Company E 
was making no further progress induced 
Olander to pull his men back to the de- 
molished enemy strongpoint and dig in 
there for the night. 

Meanwhile, Company F was undergoing 
a repetition of the trouble it had encoun- 
tered the day before. Shortly after jumping 
off, the 3d Platoon on the right discovered 
that the Japanese had mined the only 
available tank route and engineers were 
called up to abate the nuisance. Two tanks 

were then called up and succeeded in 
destroying two machine guns that lay 
athwart the line of advance. Immediately, 
another gun opened up. A squad went 
forward to take out this position but was 
pinned down by machine gun fire and a 
shower of grenades. Further tank action 
was delayed when radio communications 
between the tanks and infantry gave out, 
and not until 1500 was the platoon leader 
able to direct his tanks into the area of re- 
sistance. Finally, the two tanks succeeded 
in bringing their guns to bear against the 
position, and shortly after 1500 the whole 
platoon pushed forward and into the Jap- 
anese line. Here they found six heavy ma- 
chine guns, several mortars, a wrecked 
dual-purpose gun, and all types of grenades 
and ammunition, together with the dead 
bodies of over a hundred Japanese. The 
platoon dug in for the night. The other two 
platoons of Company F had remained 



stationary during the day guarding the 
northern approach to the valley between 
Mount Nafutan and Ridge 300. 

June 25 marked the climax of the cam- 
paign for the capture of Nafutan Point. 
During the day the 2d Battalion knocked 
out and overran the main defensive line of 
Japanese positions on top of Ridge 300. 
These positions controlled the approach to 
the point, and it was from Ridge 300 that 
the advance of the whole line had been 
held up since 22 June. 

Plans for 26 June were the same as on 
the previous day except that the I St and 
2d Platoons of Company F were to leave 
the northern mouth of the valley and take 
position on the left flank of the battalion 
line. At 0645 concentrated mortar fire 
was directed along the whole front, and at 
0750 both batteries of 90-mm. antiaircraft 
guns fired a ten-minute concentration. 
Promptly at 0800 all three companies 
jumped off. 

On the right, Company E moved slowly 
forward, fighting the terrain and the un- 
derbrush. By 1400, when it was some fifty 
yards ahead of its farthest point of advance 
of the previous day, a machine gun opened 
up directly in front of the 2d Platoon. A 
self-propelled mount came forward but 
could not bring its gun to bear against the 
enemy position. Finally, the enemy gun 
was taken out by a BAR belonging to 
Company G, whose right flank was mov- 
ing along an elevation to the left of E 
Company and was therefore in a better 
position to fire on the enemy in front of 
the latter unit. That company resumed its 
advance and for the next 200 yards met no 
opposition. At 1600 Captain Smith was no- 
tified that the other two companies were 
pulling back to approximately the same 

positions they had held the night before, 
so he did likewise. 

Company G made more rapid progress. 
After cleaning out the position to the front 
of E Company, Captain Olander's men 
pressed ahead. At 1600 their tanks left to 
return to their maintenance pool for the 
night, but the company commander elec- 
ted to go on without them. Within half 
an hour his men had arrived at the south- 
ern edge of Ridge 300. 

It was on the left flank in the zone of 
Company F that the greatest progress was 
registered on the 26th. With three platoons 
abreast, and without benefit of tank sup- 
port, the company pushed steadily forward 
without meeting any enemy fire. By 1700 
it had reached the southern end of Mount 
Nafutan, a thousand yards from the tip of 
the peninsula. There, the men began to re- 
ceive small arms fire and came to a halt. 
At 1830 F and E Companies withdrew all 
the way back to the area in which G had 
spent the previous night. This withdrawal 
was made because both company comman- 
ders felt that their positions on the top of 
the high rocky points of Mount Nafutan 
and Ridge 300 were too exposed to provide 
satisfactory spots to dig in and establish 

The battalion dug in in four perimeters 
on the night of 26 June. The three rifle 
companies, less E Company's 3d Platoon 
but reinforced by elements of H Company, 
dug in on Ridge 300. The 3d Platoon of 
E Company still occupied the old bivouac 
area near the west coast of the peninsula. 
The whole area between the 2d Battalion 
positions on Ridge 300 and the sea to the 
east was unoccupied by American troops 
and serious gaps appeared on the right of 
the line. 

Americans Rescuing Baby From Nafutan Cave 



Shortly after midnight of 26 June, a body 
of Japanese estimated at 500 sneaked 
through the 2d Battalion's outposts. Their 
destination was Hill 500, formerly the site 
of headquarters of the 47th Independent 
Mixed Brigade, but now occupied by the 
25th Marines in Northern Troops and 
Landing Force reserve. One small force hit 
the rear command post of the 2d Battalion, 
105th Infantry, and in the darkness was 
driven off with a loss of twenty-seven dead 
in return for four Americans killed and 
twenty wounded, 61 Otherwise the infiltra- 
tion was undetected. 

This desperate Japanese move was led by 
Captain Sasaki, commanding officer of the 
317th Independent Infantry Battalion, 
47th Independent Mixed Brigade. The 
troops composing the force consisted of 
those remnants of his own command that 
had escaped to Nafutan plus scattered 
Army and Navy men from other units, 
including the service and antiaircraft 
troops formerly stationed on Aslito. 

Sasaki's order read in part: 

■26 June 1944 

1. The enemy situation is the same as you 
have been informed. 

2. The Battalion will carry out an attack 
at midnight tonight. After causing confusion 
at the airfield, we will advance to Brigade 
Headquarters in the. Field, 

3. . . . Units will assemble at 1930 in 
areas to be designated scyjaratcly. You must 
carry out the attack from the designated 

4. Casualties will remain in their present 
positions and defend Nafutan Mount. Those 
who cannot participate in combat must com- 
mit suicide. 

5. We will carry the maximum of weapons 
and supplies. 

6. The pass word for tonight will be 
"Shichi Sei Hokoku" [Seven lives for one's 
country]. 62 

The word "battalion" as applied here is a 
courtesy title only. The force was a con- 
glomerate mixture of all kinds of troops, of 
which the remnants of Sasaki's battalion 
formed only the nucleus. 

About 0230 Sasaki's force hit Aslito field 
and splattered the area with machine gun 
and small arms fire before moving on 
toward Hill 500, where it apparently ex- 
pected to find the command post of the 
47th Independent Mixed Brigade. Arriving 
at its destination around 0530, one part of 
the force found instead that the hill was 
occupied by the 2gth Marines, who in- 
stantly gave battle with small arms and 
hand grenades. 

Simultaneously, another group of Japa- 
nese fell upon the 14th Marine Artillery 
Regiment in positions between Hill 500 and 
Aslito. Here another hot fight ensued, the 
Marine artillerymen killing 143 Japanese 
at the cost to themselves of 33 killed and 
wounded. 03 Still another segment hit the 
command post of the 104th Field Artillery 
Battalion, where 15 to 20 of them were 
killed. The 25th Marines mopped up the 
remaining stragglers the next morning, 
and with that Sasaki's breakthrough was 
finished. 64 

On the morning of 27 June all three 
companies of the 2d Battalion, 105th In- 
fantry, formed a skirmish line and swept to 
the end of the peninsula with no trouble. 
Not a live Japanese was encountered, and 
at 1840 Nafutan Point was declared se- 
cure. 65 Two hundred dead Japanese, most- 
ly soldiers, were found in five of the caves 

Ibid., p. 216. 

es NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, p. 34. 

63 Ibid; TF 51 Opn Rpt Marianas, Incl A, p. 10. 

ai NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, p. 34; TF 51 Opn Rpt Marianas, Incl A, p, 
10; 27th Inf Div G-a Jnl, 28 Jun 44, Msg 19. 

68 27th Inf Div G-2 Jnl, 27 Jun 44, Msg 64. 



on Mount Nafutan, and later another 350 
dead enemy soldiers were counted in the 
area of the operation of the 2d Battalion, 
105th Infantry.* 50 These figures, plus the es- 
timated 500 that had participated in the 
breakthrough, bring the total estimate of 
enemy combat personnel in the area to 
about 1,050, considerably above the origi- 
nal estimate of 300 to 500 that had been 
agreed upon by the 105th Infantry, the 
27th Infantry Division, and Northern 
Troops and Landing Force. 

Also captured on Nafutan Point on 28 
June were four 6-inch guns of British 
manufacture and three 14-cm. guns man- 
ufactured in 1925 at the Yokosuka Naval 
Arsenal. All were in the vicinity of the ra- 
dar station on the point but had not yet 
been emplaced. One of the 14-cm. guns 
was slightly damaged; one 6-inch gun was 
badly damaged, two were slightly dam- 
aged, and one was almost intact. 67 

Nafutan Point had taken a long time to 
capture, probably longer than was neces- 
sary. General Holland Smith and his staff 
were bitterly disappointed, not to say out- 
raged, by the slow progress made by the 
2d Battalion, 105th Infantry. They com- 
plained frequently about "1000 Americans 
being held up by a handful of Japs." 08 

The number of enemy troops isolated on 
Nafutan Point was actually considerably 
more than a handful, and probably totaled 
about 1,050. Also, the effective strength of 
the 2d Battalion, 105th Infantry, as of 27 
June, was down to 556 officers and men 
according to Colonel O'Connell, USA, who 

66 O'Connell, Opns in Nafutan Point Area, Sai- 
pan, a Jul 44, Buckner Board Rpt, Exhibit XX, 
pp. 2-3. 

07 Ibid. 

fi8 O'Connell Rpt, Buckner Board Rpt, Exhibit 
WW; Testimony of Maj Norvel H. Moore, USA, 
Exhibit BBB. 

commanded the unit during the last stage 
of the capture of Nafutan. 00 

In addition, the terrain that the Ameri- 
can soldiers faced was far from suitable for 
rapid assault. As described in the battalion 
report, "The terrain consisted of steep 
ridges, deep gulches with cliffs, ground 
broken with coral pinnacles, and thick 
jungle type underbrush which impeded 
progress and made observation impos- 
sible." ™ Also, for the first three days of 
the assault, the battalion had no artillery 
support, and after that only the 40-mm. 
and the 90-mm. antiaircraft guns that 
Colonel O'Connell brought down when he 
took over command, plus naval gunfire 
from three destroyers. 

The low estimate of the number of Jap- 
anese troops in the area that was enter- 
tained by corps headquarters was derived 
from an intelligence report emanating from 
the 27th Division itself. As of 21 June, the 
division had estimated the number of re- 
maining Japanese on Nafutan to be be- 
tween two and three hundred. 71 Since no 
change in this figure had been made, Gen- 
eral Holland Smith's staff had some reason 
to assume that only a "handful" remained. 
Also, the bare figure of 1,050 enemy troops 
cited above offers no real picture of the 
combat efficiency of the Japanese left on 
the peninsula. These were, it must be re- 
membered, stragglers who had made a 
disorderly retreat before the onslaught of 
the American push across Aslito field. They 
were disorganized, short of supplies, and 
in some cases unarmed. 

Against these people, the American drive 

6<J Buckner Board Rpt, Exhibit WW. 

70 2d Bn 105th Inf Narrative of Events, Saipan. 

P- 3- 

71 27th Inf Div G-2 Periodic Rpt 5, 21 Jun 44. 



was halting and slow. There was some 
justification for Holland Smith's lack of 
confidence in the leadership of the regi- 
ment, and later of the battalion, committed 
to cleaning up Nafutan. The attack of the 
infantry companies was frequently unco- 
ordinated; units repeatedly withdrew from 

advanced positions to their previous nights' 
bivouacs; they repeatedly yielded ground 
they had gained. Whatever the extenuat- 
ing circumstances, these facts could not 
fail to raise doubts about the aggressiveness 
and combat efficiency of the unit assigned 
to the mission. 


»T:ojrn I HF 1-1 



Matuton (■! 

All positions ore opproiimole 

1000 2000 



The Fight for Central Saipan — I 

Preparations for the Drive 
to the North 

While elements of the 27 th Division were 
slowly inching their way down Nafutan 
Point, the two Marine divisions prepared 
to launch the drive against the main line of 
Japanese defenses, which stretched across 
the waist of the island from just below 
Garapan to the northw est corner of Magi- 
cienne Bay. 1 (Map 6) The 2d Marine Di- 
vision in the north had little more to do 
than consolidate the lines it already held, 
send out patrols, and mop up small iso- 
lated pockets of enemy troops still lurking 
within its sector. The 4th Marine Division, 
before it could be in a position to attack, 
would have to reorient the direction of its 
drive from east to north and then push 
forward (northward) about a thousand 
yards to tie its left flank in with the right 
flank of the 2d Division. When this was 
accomplished the two Marine divisions 
would be drawn up abreast on what was 
designated the O-4 line, which ran roughly 
parallel and a little to the south of the 
Japanese main line of defense. 2 

During the period in which the 4th Ma- 
rine Division was pivoting to the left, the 
only serious fighting occurred around Hill 

500, on 20 June. This 500-foot eminence 
just west of the village of Tsutsuuran had 
once been the site of the command post of 
Colonel Oka, commander of the 4yth In- 
dependent Mixed Brigade, who had since 
left it for a safer location to the northward. 
Hill 500 fell within the zone of the 25th 
Marines, which attacked it in column of 

Following an advance preparation of 
rockets, artillery, heavy weapons, and mor- 
tars, the lead battalion moved forward 
about 1030 under cover of smoke. By noon 
it had seized the hill, and it spent the rest 
of the day mopping up the network of 
caves that ran through the area. All to- 
gether, the marines suffered forty-nine 
casualties and accounted for forty-four 
enemy dead. The hill had been well organ- 
ized for defense but not strongly manned. 3 

That same day the 8th Marines, which 
constituted the 2d Marine Division's right 
(south) flank, made a forward advance 
against no opposition to tie in with the left 
flank of the 4th Division. There was little 
other activity in the 2d Division's zone of 
action. 4 By nightfall of the 20th the ma- 
rines rested securely on the designated 
O-4 line ready to jump off on order for 
the big drive northward. They spent June 

1 See above, pp. 1 16-17. 

- See NTLF Opn Order 8-44, 19 Jun 44, Annex 
A, and Japanese Situation Map, 19 Jun 44, Incl 
to Ch. VI. 

3 25th RCT 4th Marine Div Rpt, p. 5; Hoff- 
man, Saipan, p. 1 15. 

1 Hoffman, Saipan, p. 114. 



Hill 500. Marines hi foreground await signal to advance on the hill. 

2 1 st resting and sending out patrols. Men 
of the 4th Division moved as far as 1,500 
yards to their front without meeting any 
organized enemy resistance. 5 

Landing the 106th Infantry 

Before leaving Hawaii, the 106th Infan- 
try Regiment had been assigned as reserve 
for the 27th Division with the probable 
mission of landing on Saipan. On arrival 
at Kwajalein, Col. Russell G. Ayers, the 
regimental commander, was informed that 
his unit would be attached to the Southern 
Landing Force, destined for Guam. The 
regiment was to land in the rear of the 1st 
Provisional Marine Brigade and to capture 
Orote Peninsula on that island. 6 On 16 
June, however, because of the imminent 

engagement with the approaching Japanese 
fleet, Admiral Spruance indefinitely post- 
poned the landing on Guam, 7 and on the 
1 8th the transports carrying the 106th In- 
fantry were detached from Admiral Conol- 
ly's Southern Attack Force and ordered to 
Saipan. 8 

On the evening of the 1 8th General Hol- 
land Smith requested that the 106th be 
landed "in order to maintain the continuity 
of the offensive," but Admiral Turner was 
reluctant to comply because to do so 
would inevitably delay the Guam attack 
until another reserve force for that landing 
could be brought up from the Hawaiian 
area. Nevertheless, Smith continued to 
press his case, and at last Turner con- 

D Ibid., p, 120, 

6 106th (Reinf) RCT Narrative Rpt, Forager 
Opn, 29 May-14 Aug 44, pp. 1-2. 

7 See above, p. 121. 

s 106th RCT Rpt Forager, p. 2. The other 
landing troops originally assigned to Guam were 
subsequently returned to Eniwetok. TF 53 Rpt of 
Amph Opns for Capture of Guam, 10 Aug 44, 
pp. 5-6. 



Marines on the Crest of Hill 500 mop up network of caves. 

currcd and ordered the 106th Infantry to 
commence landing early on the morning of 
20 June. 9 

The regiment landed on order and ex- 
cept for its attached artillery was assigned 
to Northern Troops and Landing Force 
reserve. Next day Colonel Ayers directed 
all of his units to initiate reconnaissance 
in the zones of action of both Marine di- 
visions. In conducting this reconnaissance, 
one group from the Antitank Company was 
ambushed and suffered four casualties, the 
first to occur in the regiment. 

The rest of the day the ist Battalion 
patrolled Susupe swamp with the mission 
of clearing Japanese stragglers from the vi- 
cinity of corps headquarters and a nearby 
Marine hospital. For this purpose the 
battalion was attached to the 2d Marine 
Division under whose control it remained 

until the morning of 23 June. During this 
period the battalion killed eighteen Japa- 
nese in the swamp and took two prisoners 
of war. 10 

Japanese Situation on the Eve 
of the Northern Attack 

While the marines pivoted on the 2d Ma- 
rine Division's left flank below Garapan, 
the Japanese too were pivoting on almost 
the same point. By the igth they were in 
position along a new "line of security" that 
ran from below Garapan, east to White 
Cliff, south to Hill 230, and then southeast 
through Hill 286 to Magiciennc Bay. 11 
General Saito disposed his troops in new 
sectors divided by Mount Tapotchau. On 
the extreme right (west) of the Japanese 
flank the town of Garapan was occupied 

Comdr Amph Forces, Ser. 000156, 4 Sep 44, 
Incl A; 1 06th RCT Rpt Forager, p. 2. 

10 1 06th RCT Jnl Forager, p. 4, 

1 1 See above, p. 116. 




Japanese Type 93 13.3-MM. Machine 
Gun. This dual-purpose gun was cap- 
lured at Salpan. 

by naval units, chiefly the Yokosuka 1st 
Special Naval Landing Force. To their left, 
the 135th Infantry held the area between 
Garapan and the west slopes of Mount 
Tapotchau, The 118th Infantry, a strag- 
gler unit, was to hold the area southeast 
of Tapotchau and be prepared to check 
enemy landings from Magiciennc Bay. Rag- 
man Peninsula was to be held by those 
remnants of the 47th Independent Mixed 
Brigade that had not already been de- 
stroyed or isolated on Nafutan Point. The 
gth Expeditionary Unit, another straggler 
force, was placed under command of the 
4jth Independent Mixed Brigade and as- 
signed to defend the shore north of Rag- 
man Peninsula. In general reserve was the 
136th Infantry, which had been ordered to 
assemble at Chacha at sunset on the 19th. 
The gth Tank Regiment had a dual 

mission — to co-operate with the 1 18th Reg- 
iment, and to check any advances along 
the coast or against the beaches of Magi- 
cienne Bay. 12 

Even this late in the campaign, the 
Japanese expected either a landing on 
Magicienne Bay or a tank attack up the 
bay coast. Their fear of American tanks 
was especially acute, as a report from 31st 
Army headquarters attests: "The changes 
in the battle up until today have been the 
results of naval gunfire and bombing but 
from now on the main thing will be to 
gain unfailing victory in antitank warfare. 
Our army has new ideas concerning this 
point and we hope this is not a miscal- 
culation." 13 

The stubborn determination of the Jap- 
anese to continue their resistance is all the 
more remarkable in view of their losses to 
date. As of 19 June approximately three 
and a half of the 43d Division's original 
eight battalions had been destroyed. Only 
one of five artillery battalions remained. 
The 4Jth Independent Mixed Brigade had 
been all but eliminated as an organized 
fighting unit. Two and a half battalions of 
infantry belonging to other units were 
destroyed, only one composite battalion 
remaining. Sixty percent of the gth Tank 
Regiment was destroyed, as were most of 
the yth and nth Independent Engineers. 14 
On the eve of the American attack to the 
north the personnel losses of Japanese line 
units were reported to be not lower than 
50 percent. 15 

In terms of artillery and tanks, the Jap- 
anese were just as badly off and as hope- 

12 Japanese Studies in World War II, 55, map 
facing p. 79. 

13 31st Army Msg File, Msg 1059; CINCPAC- 
CINGPOA Trans 9983-85. 

14 31st Army Msg File, Msg 1060. 

15 Ibid., Msg 1071. 



Japanese Field of Fire from ML Tapolchau. 

lessly outnumbered. All that remained to 
them on 20 June after six days of fighting 
were eleven 75-mm, field pieces, twenty- 
seven tanks, three operational antiaircraft 
guns, and nine machine cannon. 10 

But though the Japanese on Saipan were 
weak in manpower and short of weapons 
and equipment, they lacked nothing of the 
traditional spirit that had driven and was 
to drive so many of their countrymen to 
glorious if futile death on the battlefield. 
On the eve of the battle one tanker doubt- 

10 Corrected figures are derived from 31st Army 
Msg File, Msgs 1060, 1064, 1068. These figures do 
not include the equipment in the hands of strag- 
gler elements on Nafutan Point. 

less spoke for most of his compatriots when 
he inscribed in his diary : 

The fierce attacks of the enemy only in- 
crease our hostility. Every man is waiting for 
the assault with all weapons for close quar- 
ters fighting in readiness. We are waiting 
with 'Molotov cocktails' and hand grenades 
ready for the word to rush forward recklessly 
into the enemy ranks with our swords in 
hands. The only thing that worries me is 
what will happen to Japan after we die. 17 

22 June: The Jump-off 

General Holland Smith's orders for 22 
June called for an attack to the north by 

17 CINCPAC-CINCPOA Trans 10238, extracts 
from the diary of Matsuya Tokuzo. 



the two Marine divisions in line abreast, 
the 2d Division on the left, the 4th on the 
right. The jump-off hour was to be 0600, 
the objective line (O-5) to run through 
Laulau village on Kagman Peninsula on 
the right, Mount Tapotchau in the center, 
and a point on the west coast of the island 
about 1,000 yards south of Garapan on 
the left. 18 

In view of the fact that this northward 
push would automatically extend the lines 
of the two Marine divisions, especially as 
the 4th Division was required to spread 
eastward on Kagman Peninsula, the corps 
commander alerted the 37th Division to 
the fact that it might soon be committed 
to the northern line. The Army division, 
which was then in corps reserve, was or- 
dered to reconnoiter routes to both of the 
Marine divisions' zones of action. 37th Di- 
vision Artillery was passed to the control 
of General Harper's XXIV Corps Artillery 
to deliver close and deep support missions 
in advance of the marines. All together, the 
troops on the northern line would have 
eighteen battalions of artillery to support 
their drive on the morning of 22 June. 19 

On the right of the 4th Marine Division 
zone in the area inland from Magicienne 
Bay, the 24th Marines jumped off on 
schedule, made rapid progress against light 
opposition, and by 1330 reached an inter- 
mediate objective line that had been estab- 
lished by the division commander about 
2,000 yards in front of its line of de- 
parture. 20 

To its left, the 25th Marines found the 
going more difficult. The regiment jumped 
off on schedule in column of battalions. 

1S NTLF Opn Order 9-44, 2 1 Jun 44. 

la NTLF Rpt Marianas, Phase I, Incl E, G-3 
Periodic Rpt 8. 

3,1 4th Marine Div Opns Rpt Saipan, Annex I, 
24th RCT Rpt, p. 20. 

and within half an hour the lead battalion 
was attacked by a force of Japanese troops 
accompanied by a tank. Ninety of the 
enemy were killed and the tank was de- 
stroyed. For the rest of the morning the 
regiment encountered light resistance, but 
just south of the intermediate objective line 
received severe machine gun fire, a situa- 
tion that was aggravated by an exploding 
Japanese ammunition dump. This slowed 
progress so that by the day's end the 25th 
Marines were still short of the day's objec- 
tive, although the regiment had made an 
advance of about 2,000 yards. 21 

Meanwhile, the 23d Marines, which had 
been in division reserve, was committed 
between the two assault regiments shortly 
after noon. Fighting against light enemy 
resistance but over stubborn terrain, the 
23d Regiment, too, fell just short of reach- 
ing the day's objective by the time it dug 
in for the night. 22 

The 2d Marine Division had the more 
difficult task of gaining the approaches to 
Mount Tapotchau and of pushing to the 
top of Mount Tipo Pale, General Watson, 
the division commander, placed all three of 
his regiments in line abreast, the 8th, 6th, 
and 2d Marines from right to left. As the 
official Marine Corps historian described 
the scene, "Looking to the north of the 
6th and 8th Marines' lines, a nightmare of 
sheer cliffs and precipitous hills could be 
observed, separated in crisscross fashion by 
deep gashes .... Dense foliage which 
cloaked the region often limited visibility 
to a few feet." 2S 

Both assault battalions of the 8th Ma- 
rines made fair progress against little 

21 Ibid., Annex J, p. 6; Hoffman, Saipan^ p. 127. 

22 4th Marine Div Opns Rpt Saipan, Annex H. 
23d RCT Rpt, p. 40. 

" 2S Hoffman, Saipan, p. 128. 



opposition during the morning even 
though their axis of attack was cut by deep 
ravines, cliffs, and transverse ridges. As 
they neared the top of the first ridge line, 
they lost contact, and about noon the re- 
serve battalion had to be committed in the 
center. An hour later forward movement 
stopped as mortar fire began to fall heavily 
all along the line. Enemy machine guns lo- 
cated on a hill to the right in the zone of 
the 25th Marines commenced to lay en- 
filade fire along the right flank of the 8th 
Marines. The summit of Mount Tapotchau 
still lay about 1,200 yards (as the crow 
flies) ahead, but no further progress could 
be made that day, and the regiment dug in 
for the night. 24 

In the center the 6th Marines was ini- 
tially held back by the slow progress of the 
division's right flank. Shortly after noon 
the advance toward Tipo Pale got under 
way, only to come up against several 
pockets of Japanese machine guns, which 
stopped the lead elements on the slopes of 
the mountain. After a futile attempt to get 
at these positions by a flanking movement, 
the marines bypassed them altogether, and 
by 1400 the lead battalion had pushed to 
the top of Tipo Pale. 2 '" 

The advance to the top of Tipo Pale 
marked the farthest and most significant 
progress in the zone of the 2d Marine Di- 
vision. This eminence was about 1,000 feet 
in height and lay about 1,200 yards south- 
west of Tapotchau's summit. 2 ''' Its capture 
was essential to cover any approach to the 
western slope of Tapotchau. 

No forward movement was made on the 
extreme left of the 2d Division's zone. The 
2d Marine Regimental Combat Team oc- 
cupied the O-5 line south of Garapan for 
several days, and since the whole forward 
maneuver of the two divisions pivoted on 
this regiment it was forced to remain sta- 
tionary until the other regiments pulled 
abreast. 27 

Meanwhile, preparations were proceed- 
ing apace to move the 27th Infantry Divi- 
sion into the main line of attack. On the 
evening of 2 1 June General Holland Smith 
had ordered the division to conduct recon- 
naissance to the north over the road net 
that led to the ad and 4th Marine Division 
areas. aH By late afternoon of the 2 2d all 
three battalions of the 165th Infantry had 
completed the reconnaissance as ordered 
by Colonel Kelley. 29 That evening Colonel 
Ayers took the commanders of the 2d and 
3d Battalions, 106th Infantry, with him on 
a road reconnaissance and before dark had 
reached a point where his regiment would 
leave the road system and move cross coun- 
try in the direction of the front held by 
the 4th Marine Division. 30 

About 1600 that afternoon General 
Ralph Smith visited the headquarters of 
General Erskine, chief of staff to General 
Holland Smith, where he first received 
definite information that his division would 
be fed into the Marine front line the follow- 
ing day. The plan was for the Army divi- 
sion to relieve the left flank elements of the 
4th Marine Division so as to permit that 

21 8th Marines 2d Marine Div SAR Forager, 

P- 4 ; 

"■' 6th Marines 2d Marine Div SAR Fokagkr, 

P- 7; 

2 " Sec Map of Saipan reproduced from captured 
Japanese map by ACofS G-a NTLF, 26 J un 44, 

i7 ad Marines -id Marine Div SAR Forager, 

P- 4- 

28 NTLF Opn Order 9-44, 21 Jun 44. 

59 165th RCT Jnl, 2 3 Jun 44, Msg 15; 2d Bn 
165th RCT Jnl, 22 Jun 44, Msg 7; 3d Bn 165th 
RCT Jnl, 1% Jun 44, Msg 19. 

30 Testimony of Col Ayers, Buckner Board Rpt, 
Exhibit CCC, p. 1. 



unit to move eastward to cover Kagman 
Peninsula. This would put the 27th Di- 
vision in the center of a three division front 
and at the entrance to the valley that lay 
between Mount Tapotchau and its hill 
system on the left (west) and a series of 
hills and ridges on the right (east) that 
ran north from Magicienne Bay. 31 Jump- 
off hour for the next morning was set at 

As soon as this decision was reached, 
General Ralph Smith called Brig. Gen. 
Redmond F, Kernan, Jr., and ordered the 
division artillery to begin reconnaissance 
for positions from which to support the di- 
vision attack next morning. Smith then left 
for his own command post where he met 
with Colonels Kellcy and Ayers, com- 
manders of the two regiments that would 
go into action next day. General Smith as- 
signed Ayers' 106th Infantry to the left of 
the division line and Kelley's 165th to the 
right. Zones of approach to the line of 
departure were assigned to each of the 
regimental commanders, who in turn were 
to work out their routes of approach with- 
in their zones. 

Ayers and Kelley returned to their own 
command posts shortly after 1800 and 
began briefing their battalion command- 
ers. 32 In the 106th Regiment the 3d Bat- 
talion, under Lt. Col. Harold I. Mizony, 
was designated the assault battalion. It 
was to be followed in column by the 2d, 
under Maj. Almerin C. O'Hara, and the 
1st, under Lt. Col. Winslow Gornett. 33 
Colonel Kelley designated his 2d Bat- 

sl Information concerning this conference and 
General Ralph Smith's actions on 22 June is de- 
rived from Ralph Smith, Notes, Buckner Board 
Rpt, Exhibit M, pp. 10-13. 

32 106th RCT Jul, 22 Jun 44, p, 32. 

ss 106th Inf FO 8, 0100, 23 Jun 44. 

talion, under Lt. Col. John F. Mc- 
Donough, as the lead battalion during the 
approach. The 2d was to be followed by 
the 1st, under Major Mahoney, and the 
3d, under Major Claire. Upon relieving the 
marines, the 165th was to take up the line 
with its 2d Battalion on the left and its 
1st on the right. 34 

Orders from division headquarters con- 
firming these decisions were issued at 
2ioo. 3n The line of departure for the two 
Army regiments was to be the "line held 
by the 4th Marine Division within the 
[27th] Division zone of action." Two and 
a half hours later General Holland Smith's 
headquarters issued substantially the same 
order. 36 In this case, the line of departure 
was designated as the "front lines at King 
Hour [1000]," which was essentially no 
different from that specified by Ralph 

23 June: Into Death Valley 

Promptly at 0530, just as day was break- 
ing, both regimen ts began to move toward 
the front lines. 37 \{Map 7)] ln the 106th 
Infantry zone Colonel Mizony's 3d Bat- 
talion led off with Company L in the lead, 
followed by K, then battalion headquar- 
ters, and finally I Company. 38 Colonel 
McDonough's 2d Battalion, 165th, took the 
lead in that regiment. About 0620 the head 
of McDonough's column cut into Mizony's 
column just behind L Company, thus 
splitting the 3d Battalion, 106th Infantry, 

31 Col Kelley, Rpt of Action Saipan Island, 
165th Inf, 16-27 J un 44j Buckner Board Rpt, Ex- 
hibit N, p, 6. 

33 27th Inf Div FO 46, 222100, Jun 44. 

;ih i\| r ixjr Op n Order 10-44, 22 J un 44- 

'■" 3d Bn 1 06th RCT Jul, 23 Jun 44, p. 8; 1st 
Bn 165th RCT Jnl, 23 Jun 44, Msg a, 

:l8 3d Bn f ofith RCT Jnl, 23 Jun 44, p. 8. 


23-24 June 1944 

^^^>*. Ajis of advmc£,2J Juke 


V8iSs» in 27th Division ione 

Aft positions ore approximate 
LD as shown an NTLF G- J situation map 1800, £3 June 
Contour inftrro/ IOO feet 

-'■ /A 


Iff s 



K Tmpl, 



at that point. 30 The two regimental com- 
manders conferred but nothing could be 
done to unscramble the units until they 
reached a clearing that would permit the 
165th to move eastward and the 106th to 
proceed north toward its assigned zone of 
action. At this point of divergence an officer 
control station was set up to sort out the 
vehicles and units of the two regiments and 
direct each to its proper destination. All 
together, this delayed Companies K and I, 
1 06th Infantry, upwards of an hour, 
though Company K was due to move into 
the assault at the line of departure at 1 000. 
Company L, 106th, on the other hand, was 
ahead of the traffic jam and was able to 
push on unhindered. 40 

The two assault battalions of the 165th 
Infantry relieved the 24th Marines at 1000 
on schedule. 41 Company L, 106th Infan- 
try, completed the relief in its zone at 0930. 
Only Company K of the 106th was late, 
but its tardiness was to hold up the entire 
division attack. Not until 1055, or almost 
an hour after the scheduled jump-off time, 
was Company K in line, and not until then 
could the 3d Battalion jump off in the 
attack. 12 

When the men of the 1 06th Infantry got 
into the line, they were surprised to dis- 
cover that some of the marines whom they 
were to relieve had fallen back two or three 
hundred yards from positions held the day 
before. 451 Company K of the 25th Marines 

3fl Ibid. This movement to the line of departure 
is described in Colonel Kclley, Rpt of Action Sai- 
pan, Buckner Board Rpt, Exhibit N, and Ayers 
Testimony, Buckner Board Rpt, Exhibit GOC. 

40 27th Inf Div G-3 Jnl, 23 Jun 44, Msg 34. 

11 1st Bn 165th RCT Jnl, 23 Jun 44, Msg 4; 2d 
Bn 163th RCT Jnl, 23 Jun 44, Msg 3, 

42 106th RCT Jnl, 23 Jun 44, Msgs iro, 118. 

4:1 Ayers Testimony, Buckner Board Rpt, Exhibit 
CCC.p. 3. 

had pulled back to its right rear on the 
previous evening to tie in the night de- 
fense. 44 This caused some consternation at 
106th headquarters because the regimental 
commander was under the mistaken im- 
pression that the line of departure for the 
morning's attack was the forward line of 
the previous day's advance as indicated on 
his overlay, rather than the "front lines at 
King Hour" as stated in the operation or- 
der. Actually, of course, this withdrawal 
on the part of the Marine company eased 
Colonel Ayers' immediate problem by re- 
ducing the distance his already tardy 
troops would have to cover before reaching 
the line of departure. It did, however, 
create a gap between the 106th and the 
165th on his right that would have to be 
covered before the two units could move 
forward abreast. 

The positions the 27th Division was or- 
dered to assault had, since 19 June, been 
held by the 118th Infantry Regiment, 
which as of that date was made responsible 
for the entire Japanese line of defense from 
the east slope of Mount Tapotchau to 
Magicienne Bay. 45 The Japanese regiment 
had been torpedoed en route from Japan 
to Saipan less than three weeks earlier and 
had arrived on the island minus about 850 
of its troops and almost completely strip- 
ped of its weapons and equipment. 46 
Total troop strength of the regiment on its 
arrival on Saipan was estimated to be 
about 2,6oo. 47 The Japanese command 
had had neither time nor opportunity to 
re-equip the survivors and reorganize them 
into a first-class fighting unit. The degree 
of attrition suffered by the regiment since 

44 Hoffman, Saipan, p. 134. 
10 See above, pp. 1 16-17. 
16 See above, p. 60. 
47 See App. A. 



the American invasion is unknown but it 
cannot have escaped damage from the 
terrible pounding from air, sea, and land to 
which the island had been subjected since 
1 2 June, 

Before the 27th Division was committed, 
the 136th Infantry Regiment, 43d Division, 
which had previously been in reserve 
around Chacha village was ordered to 
move out to Hill 286 (meters), Hill 343 
(meters), "and the hills E[ast] of there." 48 
Hill 286 was in the zone of action of the 
2d Marine Division, but Hill 343, the 
"hills East," and the valley in between were 
directly athwart the line of advance of the 
27th Infantry Division. 

Initially, the 136th had been one of the 
Saipan garrison's best fighting forces, being 
at full strength and fully equipped at the 
time of the landing. However, the regiment 
had taken a frightful beating in the first 
days of the invasion. Manning the Central 
Sector facing Red and Green Beaches, it 
had borne the brunt of some of the hardest 
fighting on Saipan. Although not literally 
decimated, its combat strength had been 
severely weakened. Two men of the regi~ 
ment who had been captured on 25 June 
testified that its 2d Battalion had been de- 
pleted approximately 67 percent on the 
first day of the landing, that the remnants 
of the 2d and 3d Battalions were combined 
as a single battalion, and that the total 
strength of the regiment was less than 


But whatever losses in manpower and 

equipment the Japanese in this sector may 

have suffered, they still had one enormous 

advantage. That was terrain. 

The soldiers of the 27th Division were 
soon to dub the area "Death Valley." so 
The "valley" is really a terracelike depres- 
sion on the eastern slope of the sprawling 
mountain mass that fills most of central 
Saipan and culminates in the towering 
peak of Mount Tapotchau. The floor of 
the valley, less than 1,000 yards in width, 
is dominated along its entire length by the 
rugged slopes of Mount Tapotchau on the 
west and a series of hills, the highest about 
1 50 feet above the valley floor, on the east. 
This eastern hill system was to be called 
"Purple Heart Ridge" by the soldiers who 
fought there. Death Valley, then, was a 
sort of trough into which the men of the 
27th Division were to advance. The valley 
itself was almost devoid of cover except 
for a line of trees near the southern end 
and for three or four small groups of farm 
buildings surrounded by trees. The cliff on 
the left was for the most part bare, but 
above the cliff was wooded ground. The 
hills on the right were tree-covered. A nar- 
row road — little more than a cowpath-- 
ran up the valley a short distance then 
branched off, the left branch skirting the 
cliffs of Mount Tapotchau, the right head- 
ing toward the north face of Hill Able and 
then cutting to the east. 

Obviously, this terrain was ideally suited 
for defense against any attack through the 
valley, and the Japanese made the most of 

4K CINGPAG-GINGFOA Trans 10531, excerpts 
from a notebook of field orders, 14 Jun to Jul 44. 

49 NTLF Rpt Marianas, Phase I, Annex D, G-2 
Rpt, p. 85. 

sn This description of Death Valley is derived 
from: Brig Gen Oftden J. Ross, Summary of Opns 
by 37th Inf Div, 13 Jul 44, Buckner Board Rpt, 
Exhibit RR, p. 4; Col Albert K. Stebbins, Jr., 
Narrative Account of Opns of 27th Inf, 14 Jul 44, 
Buckner Board Rpt, Exhibit TT, p. 3 ; Sworn 
Statement of General Ralph Smith, Buckner 
Board Rpt, Exhibit AAA, p. 4; Love, The 27th 
Infantry Division in World War II, pp. 228-29. 



Hell's Pocket Area 

Treeline Deafh Valley 

Mt. Tapotchau Dominating Death Valley, where 27th Division troops fought 
horn 28 through 30 June. 

it. In the words of Col. Albert K. Stebbins, 
Jr., 27th Division chief of staff: 

The cliffs and hillsides were pocketed with 
small caves and large caves. The wooded 
area was rough, filled with boulders, and ex- 
cellent for defensive operations. Bands of fire 
were laid by the enemy thru the underbrush 
and in such manner as to make it most dif- 
ficult to discover their locations: Well- 
placed, hostile guns fired only when lines 
passed and striking our forces in the rear 
disrupted the attack. 51 

The Japanese had at their disposal all 
kinds of automatic weapons, light and 
heavy mortars, and some 75-mm. mountain 

51 Stebbins, Account of 27th Inf Opns, Buckner 
Board Rpt, Exhibit TT, p. 3. 

guns. These were well concealed, usually 
in caves whose mouths were covered with 
brush. Troops approaching through the 
valley could get at the positions only by 
direct shots from tanks or self-propelled 
guns. It was impossible to reach them with 
artillery, at least during the initial stages 
of the attack, because the axis of the 
caves was at right angles to the line of fire 
of the artillery. 

"Cannon to right of them, Cannon to 
left of them." Not so much cannon per- 
haps, but enough other fire to make this 
seem to the men caught in the middle a 
true replica of Tennyson's "Valley of 



Tree Line in Death Valley, where elements of the 2d Battalion, 165th, were 
pinned down on 23 June. 

The units of the 27 th Division lined up 
from right to left (east to west) as follows: 

165-C: 1 st Lt. Edward L. Cloyd, Jr. 

165-A: Capt. Laurence J. O'Brien 

i6f,-G: Capt, Paul J. Chasmar 

165-F: Capt. Francis P. Leonard 

106-K: Capt. William T. Heminway 

106-L: Capt. Charles N. Halldcn 

The regimental boundary line coincided 
with the road that ran through Death 

On the division right the first obstacle 
to be overcome by the 1st Battalion, 165th, 
was Hill Love, lying roughly on the border 
line between Companies A and C. This 
eminence rose about 700 feet above sea 
level, was tree-covered, and was infested 

with Japanese. A patrol from A Company 
met heavy machine gun and rifle fire that 
killed the patrol leader and wounded one 
other man. It was then decided that the 
two companies would circle the hill and 
meet at its northern base. Company C on 
the right jumped off at 1015 and by 1400 
had succeeded in working its way around 
the hill to the northern face. Here, the men 
were pinned down by heavy fire to their 
front and made no further advance. During 
the course of the afternoon the com- 
pany suffered three men killed and four- 
teen wounded. A platoon of tanks from 
the 762d Provisional Tank Battalion was 
brought forward in an effort to reduce the 



Driven to Concealment by the Intensity of Enemy Fire in Death Valley 
on 23 June, zjth Division soldiers worked their way into a small wooded area. 

enemy positions. One of these, commanded 
by i st Lt. Louis W. Fleck, was set on fire 
with a "Molotov Cocktail." All the tankers 
but one were killed as they emerged from 
the turret/ 2 The men of Company C who 
witnessed the incident were helpless to 
avert it because by now marines had 
pushed ahead directly into their line of fire 
and the battalion commander had ordered 
them not to fire. 

Meanwhile, Company A had also 
reached the northern face of Hill Love, It 
too could make no further progress. At 
1630 the tanks withdrew for the night and 

the two companies dug in. Company B 
was brought in and completed the encircle- 
ment of the promontory by digging in on 
the south face. 53 

The 2d Battalion, 165th Infantry, also 
had considerable trouble during the after- 
noon. After reaching the tree line that lay 
about 300 yards to their front, both of Mc- 
Donough's companies remained stationary 
for two hours, waiting for Company K, 
106th Infantry, to work its way up on the 
left. General Ralph Smith finally ordered 
McDonough to advance without regard to 

52 Appleman, Army Tanks in the Battle for 
Saipan, pp. 34-37. 

55 Love, Battle for Saipan, pp. 340, 245; Kelley 
Testimony, Buckner Board Rpt, Exhibit PP, pp. 



what was going on in the 106th area and 
instructed Colonel Kcllcy to commit his re- 
serve if necessary. 54 Regiment therefore 
ordered an attack for 1 3 i5- r>r * Company E 
was brought up and ordered to deploy one 
platoon on the regimental left to maintain 
contact with K Company of the 106th and 
the other on the right to maintain contact 
with the i st Battalion, 165th. 06 

As soon as McDonough's battalion 
moved out from the line of trees it was 
greeted by a hail of small arms, machine 
gun, and heavy weapons fire from the cliff 
line on its left. A similar concentration of 
fire from Purple Heart Ridge on the right 
soon followed and the advance platoons of 
Company F were badly hit in the cross fire. 
Under cover of a smoke screen laid down 
by the chemical battalion, the men eventu- 
ally pulled back to the tree line from which 
they had started. Company G, witnessing 
the results of F Company's advance also 
withdrew to the tree line and remained 
there for the rest of the day. 57 

In the zone of the 106th Infantry on the 
division left the action can be character- 
ized as two separate battles since K and 
L Companies had no physical contact dur- 
ing most of the day. Company K on the 
right pushed off at 1055. Shortly thereafter 
the leading scout of the advance platoon 
was struck by machine gun fire. The rest 
of the company hit the ground and was 
immediately subjected to mortar fire. The 
company commander, Captain Heminway, 
ordered his men to move forward by in- 
filtration. This movement began about 
1300, and by 1500 the company had 
worked its way into a small wooded area 

that provided some cover against the en- 
emy weapons. There it waited for L Com- 
pany on its left to come up, and since the 
latter unit did not pull abreast until 16 15 
both decided to dig in there for the 
night. 58 

Company L had arrived on the line in 
sufficient time to push the attack at 1000 
but had been held up by K Company's 
tardiness in relieving the marines in its 
zone. On Captain Hallden's left was the 
cliff of Mount Tapotchau and on his right 
a series of ravines. About 400 yards ahead, 
the cliff line receded to form a little cove 
in the mountain wall that the soldiers 
dubbed "Hell's Pocket." In the midst of 
this cove was a lone rock that rose a hun- 
dred feet and was covered with ivy. Caves 
in the rock and in the cliff walls that sur- 
rounded it provided ideal spots for Japa- 
nese machine guns. 59 

Company L advanced about fifty yards 
from its line of departure, and Japanese 
mortar fire began to fall in the area. Hall- 
den pushed his men on, moving along the 
base of the cliff, which formed the west wall 
of Hell's Pocket. As the troops probed 
deeper into the pocket an enemy mortar 
shell set off a Japanese ammunition dump 
and the flying debris kept the men pinned 
down for over an hour. Self-propelled 
mounts were brought forward in an effort 
to knock out the cave positions of the 
Japanese but the vehicles were too exposed 
to fire from above to accomplish much. 
Finally, a platoon of medium tanks came in 
to support the infantry.* 10 

By this time the gap between K and L 
Companies had grown wider so Hallden 

si 165th R.CT Jnl, 23 Jun 44, Msg 29. 

05 Ibid., Msg 13. 

50 27th Inf Div G-3 Jnl, 23 Jun 44, Msg 89. 

57 Love, Battle for Saipan pp. 237-53. 

J 58 Ibid,, pp. 253-55. 

19 Love, The syth Infantry Division in World 
World War II, p, 23 I. 

00 1 06th RCT Jnl FoRAGEft, 23 Jun 44, Msg 193. 



shifted to the right rear out of Hell's 
Pocket and by 1530 had re-established 
contact with Company K on the valley 
floor where both companies dug in for the 
night. Company I, which had been in re- 
serve most of the afternoon, was brought 
up and dug in on the rear of this posi- 
tion. 61 Progress for the day in the zone of 
the 1 06th was about 1 00 yards. 

Throughout the day the 106th Regiment 
experienced considerable difficulty in main- 
taining contact with the 2d Marine Divi- 
sion on its left. The corps order had 
stipulated that the burden of contact was 
from right to left. 62 Responsibility for 
contact therefore rested with L Company, 
106th, but the marines were moving along 
the top of the cliff at whose base the Army 
troops were located, and physical contact 
was impossible. As Company L moved to 
the right to tie in with Company K, even 
visual contact was lost. At 1703 division 
headquarters ordered Colonel Ayers to gain 
contact with the marines on his left "with 
sufficient force to maintain it," e3 and 
half an hour later Ayers ordered the 2d 
Battalion, 106th, to cover the gap. 64 On 
the theory that it would be easier to main- 
tain contact between two companies of the 
Army division than between Marine and 
Army units, Company F was ordered to 
move to the top of the cliff and tie in with 
the marines while G Company moved for- 
ward to positions below the cliff and 
established contact on the left of L Com- 
pany. This move was not completed until 
19 10, and both companies had to build up 
a defensive line under cover of darkness." 5 

,!1 Love, Battle for Saipan, pp. 255^58. 
ai NTLF Opn Order 10-44, 22 jun 44. 

63 27th Inf Div G-3 Jnl, 23 Jun 44, Msg 126. 

64 Ibid., Msg 134. 

65 ad Bn to6th RCT Jnl, 23 Jun 44, Msg 85; 
106th RCT Jnl, 23 Jun 44, Msg 201; Hoffman, 
Saipan, p. 137. 

Later that night, Company F's ist Platoon, 
which was hugging the cliff overlooking the 
edge of Hell's Pocket, was attacked by a 
party of about fifteen Japanese who made 
their way through the perimeter before 
being discovered. In the intense hand-to- 
hand fight that ensued, bayonets, grenades, 
knives, and fists came into play. The Jap- 
anese killed two men of the platoon and 
wounded two others before being destroyed 
or routed. 66 

Progress in the zone of the 27th Division 
on 23 June had been disappointing, espe- 
cially in the area assigned to the 106th 
Infantry. During the afternoon General 
Holland Smith expressed his alarm over the 
situation in a conversation with General 
Jarman, the island commander and the 
senior Army officer on Saipan. In General 
Jarman's words: 

General Smith, CG of the V Phib Corps, 
called me to his quarters and indicated that 
he was very much concerned about the sit- 
uation which he was presented with in re- 
gard to the 27th Div. He outlined to me the 
many things that had happened with respect 
to the failure of the 27th Div to advance. 
He indicated that this division had suffered 
scarcely no [sic] casualties and in his opin- 
ion he didn't think they would fight. . . . He 
stated that if it was not an Army division 
and there would be a great cry set up more 
or less of a political nature, he would im- 
mediately relieve the division commander 
and assign someone else. 67 

Next morning, General Holland Smith 
registered his displeasure in a stern dis- 
patch to General Ralph Smith himself: 


nr ' Love, Battle for Saipan, p. 362. 
r ' 7 Buckner Board Rpt, Exhibit J. 




There can be no doubt of the truth of 
General Holland Smith's charges that the 
27th Division had been late in the jump- 
off, that its advance had been slow, and that 
it had held up progress of the two Marine 
divisions on its flanks. It is apparent, how- 
ever, that he underestimated the stubborn- 
ness of the Japanese defenses in the area 
by dismissing the opposition in the zone 
of the 27th Division as being "only by 
small arms and mortar fire." 

When subsequently queried on this 
point, Colonel Ayers, commanding officer 
of the 1 06th Infantry, was of the firm be- 
lief that if he had tried to advance rapidly 
across the open ground in front of him 
his regiment "would have disappeared." 69 
General Ralph Smith agreed. He later testi- 
fied that, after visiting the front lines 
shortly after noon, he was "satisfied that 
Col. Ayers was making every effort possible 
to advance in the valley, and considered 
that any further pushing of troops in that 
zone would only lead to increased casual- 
ties, without accomplishing adequate re- 
sults." 70 

On the other hand, General Jarman 
testified that in his conversations with 
General Ralph Smith on the afternoon of 

,iS Ibid., Exhibit G. 

,!9 Ibid., Ayers Testimony, Exhibit CCG, p. 5. 
70 Ibid., Ralph Smith, Sworn Statement, Exhibit 
AAA, p. 5. 

23 June the division commander had been 
far from satisfied with the conduct of his 
troops. In General Jarman's words: 

I talked to General Smith and explained 
the situation as I saw it and that I felt from 
reports from the Corps Commander that his 
division was not carrying its full share. He 
immediately replied that such was true; that 
he was in no way satisfied with what his 
regimental commanders had done during the 
day and that he had been with them and 
had pointed out to them the situation. He 
further indicated to me that he was going to 
be present tomorrow, 24 June with his di- 
vision when it made its jump-off and he 
would personally see to it that the division 
went forward .... He appreciated the sit- 
uation and thanked me for coming to see 
him and stated that if he didn't take, his 
division forward tomorrow he should be re- 
lieved. 71 

The First Night in Death Valley 

At 1925, just as darkness fell, the Japa- 
nese launched a six-tank attack down the 
road that ran the length of Death Val- 
ley and marked the boundary between 
Mizony's and McDonough's positions. 72 
Not until the column had almost reached 
the American outposts was it discovered, 
and by then the lead tank was too close to 
be fired upon from cither side of the road 
without endangering the men on the other. 
The other five tanks, however, were taken 
under fire by both battalions with every 
weapon available. Bazookas, antitank guns, 
grenade launchers, and artillery went into 
action, and all five tanks were knocked out. 
The lead tank proceeded on through the 
lines and circled back, firing constantly. 
One shell landed in a Japanese ammuni- 
tion dump located in the midst of the 3d 

71 Ibid., Exhibit J. 

7 - 1 06th RCT Jnl, 23 Jun 44, Msg 202. 



Marines Emerging From Purple Heart Ridge Complex 

Battalion, io6th's, lines and set it afire. 
The tank then turned east and was finally 
knocked out in the zone of the 23d 
Marines. 73 

Meanwhile, the ammunition dump in 
the middle of the 3d Battalion, 106th, was 
going off in all directions. Simultaneously, 
the Japanese on Mount Tapotchau began 
to throw mortar shells and machine gun 
fire into the area. Company L suffered 
sixteen men wounded within the space of 
an hour. The position of the entire 3d Bat- 
talion was now untenable. 1st Lt. George 
T. Johnson of I Company ordered his men 
to disperse as soon as the dump started to 

explode. He later assembled them across 
the road to the rear of the 165th line and 
dug in there for the rest of the night. The 
other two infantry companies withdrew 
about 100 yards behind the conflagration, 
thus canceling altogether the small gain 
made by the battalion during the day's 
action, 74 

23 June: Marines on the Flanks 

On the right of the 27th Division, the 
4th Marine Division attacked with two 
regiments abreast, 24th on the right and 
23d on the left. The 24th Marines pushed 

73 Love, Battle for Saipan, pp. 264-65. 

Ibid., pp. 266-67. 



rapidly ahead along the shore of Magi- 
cicnnc Bay and by midafternoon had 
reached the O— 5 line at one point just cast 
of the village of Laulau. 75 On the left the 
23d Marines made somewhat slower prog- 
ress, partly because its advance was held 
back by the 1 65th Infantry. Within a short 
time after the jump-off, one battalion 
seized the top of Hill 600, which though 
lightly manned by the enemy was admir- 
ably suited for defense and took thirty 
minutes of close fighting to capture. There, 
the battalion was ordered to hold, pending 
the advance of the Army troops on the left, 
but since the 1st Battalion, 165th, was held 
up, the marines spent the rest of the after- 
noon in a stationary position, firing and 
pitching grenades at the Japanese who still 
occupied in force the northern face of the 
hill. 76 That night a group of enemy tanks 
launched an attack against Hill 600 but 
was repulsed with the loss of three of its 
five vehicles. 77 

On the other side of Death Valley, the 
8th Marines jumped off on schedule except 
on the right flank, which was held up by 
the late arrival of the 106th Infantry. To 
fill the gap between these units, the reserve 
battalion was ordered into position to pro- 
tect the right flank and the three battalions 
in the assault moved forward. By midafter- 
noon the right battalion seized the cliff that 
dominated the only feasible route to the 
top of Mount Tapotchau. On the regi- 
mental left, the marines ran into a nest 
of about thirty Japanese riflemen and six 
heavy machine guns, which held up their 
progress for the rest of the day. 

Soon after the 6th Marines in the center 
of the 2d Division's lina launched its at- 
tack, the right flank battalion was pinched 
out by the reduced frontage. On the left, 
the regiment made no advance during the 
day because it was already so far forward 
that any further move would have caused 
too much of a contact strain. The same 
was true of the 2d Marines on the division 
left flank. Not until the center of the corps 
line made more significant progress would 
it be safe for the elements on the left to 
move ahead. 78 

24 June: Action of the 2jth Division 

General Holland Smith's order for June 
24th called for a continuation of the at- 
tack with three divisions abreast in the 
same order as before, commencing at 0800. 
Corps artillery was assigned to general sup- 
port and ordered to reinforce the fires of 
divisional artillery. 79 In the zone of the 
27th Division, its own organic artillery 
would fire a ten-minute preparation before 
the jump-off and thereafter support the 
attack on call. 80 

165th Infantry Attack Against 
Purple Heart Ridge 

At 0705 corps reminded 27 th Division 
that the slope on the right side of Purple 
Heart Ridge was in the zone of action of 
the 165th Infantry, which would capture 
it with the help of fire from the 4th Marine 
Division on its right. 81 This meant in ef- 
fect that the 106th Infantry alone would 

75 4th Marine Div Opns Rpt Saipan, Annex I, 
24th RCT Rpt, p. 20. 

7B Ibid., Annex H, 23d RCT Rpt, pp. 40-42; 
Hoffman, Saipan, p. 136. 

11 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RCT Rpt, p. 41; Hoffman, Saipan, p. 140. 

7S 2d Marine Div SAR, Phase I, Forager, Sec. 
VI, pp. 10-11; Hoffman, Saipan, pp. ih7 _ 39- 

79 NTLF Opn Order 1 1-44, 1 100 23 Jun 44. 

80 27th Inf Div FO 47, 232100, Jun 44, 27th Inf 
Div G-3 Jnl, Incl. 

S1 27th Inf Div G-3 Jnl, 24 Jun 44, Msg 8. 



be responsible for the frontal attack up 
Death Valley. Simultaneously, the 2d Bat- 
talion, 165th, would attack along the crest 
of Purple Heart Ridge itself and the 1st 
Battalion, 165th, would move up on the 
right ( east ) of the ridge. 

Purple Heart Ridge was in reality a 
scries of hills connected by a ridge line 
running in a northerly direction. From 
south to north these hills were designated 
Queen, Love, George-How, Xray-Yoke, 
Oboe, King, and Able. 82 Hill Queen had 
already been overrun by the 4th Marine 
Division in its advance eastward toward 
Kagman Peninsula. Hill Love had been 
surrounded by the 1st Battalion, 165th In- 
fantry, which had dug in the night before 
around its base. 83 

There were obvious tactical advantages 
to an early capture of this ridge line or 
any considerable part of it. It overlooked 
Death Valley from the east just as the 
higher cliffs around Mount Tapotchau did 
from the west. The commanders were be- 
coming aware that any movement north 
through the valley could be easily inter- 
dicted by fire from the elevations on either 
side. If the lower of the two walls of the 
corridor — Purple Heart Ridge — could be 
seized, then fire could be brought directly 
to bear against the caves that dotted the 
west cliffs from which such effective fire 

8U According to standard usage in the Central 
Pacific, the maps used in the Marianas were over- 
laid with a grid. Each large square of the grid 
measured rOOO by iOOO yards and was given a 
numerical designation. These were in turn sub- 
divided into twenty-five smaller squares of 200 
square yards, each given a letter designation ac- 
cording to the phonetic alphabet then in use. 
Thus, Mill Love derived its name from the fact 
that it was located in small square "L" of large 
square "175." The squares, both large and small, 
were called "target areas," 

8S See above, p. 1 76. 

was being trained on the troops trying to 
advance through the valley below. 

At 0800 the two assault battalions of the 
165th Infantry jumped off on schedule 84 
in the same order they were in at the end 
of the previous day's fighting: On the front 
line from right to left were Cloyd's Com- 
pany C, O'Brien's Company A, Chasmar's 
Company G, and Leonard's Company F. 
In reserve for the ist Battalion was Gil's 
B Company; for the 2d Battalion, Ryan's 
E Company. The regimental plan called 
for Company C on the right flank to 
swing well to the right and then northeast 
into the zone of the 23d Marines. There it 
would wait until the rest of the battalion 
came abreast. Company B was to come up 
from behind Hill Love on A Company's 
left, execute a turning movement, and then 
attack eastward across A's front in order 
to enfilade the enemy positions to the north 
of Hill Love. 85 

Within a little more than an hour after 
the jump-off, Company C had moved to 
the right and gained physical contact with 
the marines.* 6 Lieutenant Cloyd stopped 
his advance on the northern nose of Hill 
600 to await the approach of the rest of 
the battalion on his left. He made no fur- 
ther progress that day. 87 

On the battalion left, Companies A and 
B ran into immediate trouble. As A pushed 
off from Hill Love toward the same posi- 
tions that had caused so much trouble the 
day before, it came under heavy fire from 
small arms and machine guns. Company B 
then moved up to the left and, as planned, 

S4 37th Inf Div G-n Jnl, 24 Jun 44, Msg 16. 

*° Love, Battle for Saipan, p. 277. 

8ti 27th Inf Div G-3 Jnl, 24 Jun 44, Msg 22. 

87 Unless otherwise indicated, the following ac- 
count of the actions of the 165th Infantry is de- 
rived from Love, Battle for Saipan, pp. 276-87, 



tried to execute a turning movement to the 
right across A Company's front. Company 
B, too, met a fusillade of fire and finally 
withdrew after losing twelve wounded, two 
mortally. It was only 0930, but the 1st Bat- 
talion attempted no further movement 
until midaftcrnoon. 

Colonel Kclley then decided to commit 
his 3d Battalion to the right of the line. 
The 3d was to move to the right, following 
the same route used by Company C and, 
upon reaching the latter's positions on the 
left flank of the Marine lines, pass through 
and launch an attack that would carry it 
to the top of Hill King, almost at the 
northern edge of Purple Heart Ridge. 
While this was taking place, B Company 
was to circle the pocket of resistance north 
of Hill Love and build up a line facing the 
pocket from the north. Thereafter, the 1st 
Battalion was to mop up this area of re- 
sistance, and retire into regimental reserve. 
The change of plans was agreed upon at 
0904, and orders were accordingly issued 
at ioi5- K8 The 3d Battalion completed its 
move by 1335, and by that time Company 
B had encircled the pocket and was facing 
south ready to attack. 

Meanwhile, the 2d Battalion, 165th In- 
fantry, moved off on schedule in an effort 
to take the ridge by a direct assault from 
the southwest. The night before, the bat- 
talion had occupied positions about three 
hundred yards to the north and west of 
Hill Love. G Company moved off with 
Company F following, echeloned to the left 
rear. By 1000 the lead platoon had reached 
the southern lace of Hill Xray-Yokc, about 
the midpoint of the ridge. Here the men 
found a steep gulch directly across the 

path of advance. In order to get up on the 
hill itself, they would have to climb down 
the near cliff, cross the bottom of the can- 
yon, and scale the cliff on the other side. 
As the first platoon attempted this feat of 
acrobatics it came under rifle fire from the 
hill ahead and was stopped in its tracks. 
Company G made no further progress that 

F Company to the left rear had also run 
into trouble. As it came abreast of G, Com- 
pany F spread out to the left where the 
gully in front of Hill Xray-Yoke was not 
quite so precipitous. Here the terrain was 
friendlier, but the enemy was not, and 
heavy mortar and machine gun fire kept 
the bulk of the company pinned down for 
two hours. Around 1 1 30 Captain Leonard 
sent a patrol out to his left toward a small 
house located on the floor of Death Valley. 
His purpose was to protect his flank from 
a possible attack from that direction. Of 
the twenty men dispatched on this mission 
six were wounded and one was killed by 
machine gun fire before the patrol was 
withdrawn. F Company, too, now held its 
lines for the remainder of the day. At 
nightfall the whole battalion dug in just 
below the gulch. The day's action repre- 
sented an advance of about 150 yards. 

While the stalemate was developing on 
the left flank of the regimental line, stub- 
born resistance continued on the extreme 
right. The 3d Battalion had swung right 
behind the ridge and passed through Com- 
pany C at i335- S9 Shortly after 1600 Capt. 
Howard Betts, who commanded Company 
K, made a frontal assault in column of 
platoons up Hill Xray-Yokc from the east. 
The lead platoon reached the tree line well 
up the hill without much trouble, but just 

!is 1 st Bh 16.1th RCT Jnl, 24 Jun 44, Msg 13; 
3d Rn 165th RCT Jnl, 24 Jun 44, Msg 10. 

3d Bn 165th RCT Jnl, 24 Jun 44. 



as it started into the undergrowth, Japa- 
nese machine guns opened fire from the 
left front, traversing the length of Betts' 
line and wounding two men. The platoon 
took to earth and was pinned there by a 
continuous grazing fire for almost two 
hours. Betts immediately swung his second 
platoon into line on the right, but after 
about fifty yards further advance up the 
hill, it too ran into machine gun fire and 
was stopped. By this time darkness was 
coming on, and Company K was pulled 
down the hill to tie in with I Company for 
the night. Meanwhile, the ist Battalion's 
effort to mop up the pocket of resistance 
north of Hill Love had failed, and that 
unit, with the exception of Company C, 
dug in around the same positions it had 
occupied the night before. 

106th Infantry: Into Death 
Valley Again 

The units, from right to left, on the front 
line facing the mouth of Death Valley on 
the morning of 24 June were: Company K, 
1 06th Infantry, Captain Heminway; Com- 
pany L, 1 06th Infantry, Captain Hallden; 
and Company G, 106th Infantry, Capt. 
David B. Tarrant. Company F of the 1 06th 
was still on top of the cliff, and, since its 
movement was geared to that of the 2d 
Marine Division, its actions must be con- 
sidered as separate from those of the rest 
of the 106th Regiment. 

The 3d Battalion jumped off on time, 
but immediately encountered such heavy 
mortar fire that many of the men fell back 
to the line of departure and in some cases 
behind it. 90 By 0945 the front-line troops 
had advanced from 50 to 100 yards into 

the valley, but there was no sign of any 
abatement of enemy fire, especially from 
the cliffs of Tapotchau. Division headquar- 
ters was severely disappointed, and at 1 1 2 
General Ralph Smith radioed Colonel 
Ayers: "Advance of 50 yards in 1-1/2 
hours is most unsatisfactory. Start moving 
at once." fll 

In response to this pressure, Company 
K, supported by a platoon of medium 
tanks of Company B, 763d Tank Battalion, 
immediately pushed forward into the val- 
ley. Captain Heminway had two platoons 
abreast, the ist on the right, the 3d on the 
left. He left the 2d Platoon at the entrance 
to the valley to deliver covering fire to his 
front. He also set up his machine guns on 
the high ground beside the valley road and 
had some support from M Company's 
heavy weapons. Heminway's men advanced 
in a long, thin skirmish line, moving rapidly 
toward the center of the valley. They had 
pushed forward fifty yards without event 
when the entire cliff on the left of the val- 
ley seemed to open up. The company broke 
into a run toward a fold in the ground that 
offered some cover. Here the company 
commander stopped to reorganize his line. 
Just as he got up to wave his men forward 
again he was shot in the head and killed. 
This paralyzed the entire line until ist Lt. 
Jefferson Noakes, the company executive 
officer, could come forward and take com- 
mand. 92 

Meanwhile, the platoon of tanks at- 
tached to Company K had been roaming 
around the floor of the valley trying to 
silence the Japanese fire from the cliff. One 
of the tanks, commanded by the platoon 
leader, 2d Lt. Richard Hitchner, received 

1 06th RCT Jul, 24 Jun 44, Msg 230. 

91 Ibid., Msg 243. 

92 Love, Battle for Saipan, pp. 294-95. 



Tank-Infantry Co-operation, Medium tank leading men of Company K, 106th 
Infantry, into Death Valley on 24 June, 

two hits from enemy shells, was knocked 
out, and had to be abandoned. 93 Around 
1 1 00 another tank was hit, and the whole 
platoon withdrew. 

At this juncture 3d Battalion called the 
company headquarters and announced that 
a smoke screen would be laid down and 
that under its cover the men of Company 
K were to withdraw from their exposed 
positions. Some of the men did withdraw 
under the smoke, but one platoon failed 
to get the word and remained holed up 
until two tanks that had been supporting 
L Company went over and covered its 
withdrawal. 94 

Meanwhile, Company I, under Lieuten- 
ant Johnson, had been ordered into the line 
between the other two companies of the 

,J:| Appleman, Army Tanks in the Battle for Sai- 
pan, pp. 44-46. 

94 Love, Battle for Saipan, pp. 296-98. 

3d Battalion. Reporting at ioi5, 35 John- 
son was told to move two platoons out 
into the valley in column, swing left until 
he made contact with L Company's right 
flank, and then deploy his men to the right 
to close the gap that existed between Com- 
panies L and K. Immediately upon enter- 
ing the pass into the valley, both platoons 
were subjected to brutal fire from the cliffs 
on the left and stayed pinned down for 
better than two hours. The men lay on an 
open slope without any means of protec- 
tion, unable even to lift their heads to 
fire back. Seven were killed and thirteen 
wounded. A little before 1200, as the men 
still lay out on the open ground, three Jap- 
anese tanks came down the valley road 
from the north firing in all directions. Just 
as they were about to overrun the area 

1 06th ROT Jnl, 24 .pin 44, Msg 245. 



occupied by Company I, all three were 
knocked out by antitank guns from both 
sides of the valley. 

Meanwhile, Lieutenant Johnson was 
frantically trying to get at the source of 
the enemy fire that was proving so deadly 
to his unit. He had the antitank guns of 
battalion headquarters open up on the face 
of the western cliff with canister and set 
up all of M Company's heavy machine 
guns to deliver covering fire in that direc- 
tion. His 3d Platoon, which was just at the 
entrance to the valley, took up positions 
and delivered supporting small arms fire all 
along the face of Mount Tapotchau. John- 
son himself took his company headquarters 
and moved along the edge of the trees to 
the left in an effort to get at the Japanese 
position in front of Company L. As this 
group moved to the left through the un- 
dergrowth, it too was pinned down by mor- 
tar fire and discovered the woods to be 
full of enemy riflemen between the valley 
road and L Company's right flank. John- 
son was wounded. 

It was at this point that the battalion 
commander ordered K and I Companies 
to withdraw under cover of smoke. In the 
latter unit's zone, the screen was neither 
effective nor of long enough duration, and 
only a handful of Johnson's men could get 
out. Finally, about 1225, Colonel Mizony 
brought every weapon he had to bear on 
the Japanese positions along the cliff and 
in Hell's Pocket. Under cover of this fire 
the remainder of Company I was able to 
crawl and scramble back to the cover of 
the trees. 9fi Casualties in the 3d Battalion, 
1 06th, alone, for the day had amounted to 
14 killed and 109 wounded. 97 

Meanwhile, the 1 st Battalion, 106th In- 
fantry, had been ordered up to relieve the 
3d Battalion, The relief was accomplished 
at 1 5 15.°° Upon assuming its position in 
the line the 1st Battalion was ordered to 
dig in for the night since it was considered 
too late in the day to launch further at- 
tacks into the valley." 

While the 3d Battalion had been cn- 
cngaged in trying to push out into the val- 
ley, Company G, on the left flank, had 
advanced no more than two hundred yards 
during the day. The unit had not itself en- 
countered much enemy fire, but when 
Company L ran into so much trouble dur- 
ing the morning, Captain Tarrant held his 
line firm rather than push out ahead of the 
unit on his right. 100 

The attempt of the 106th Infantry to 
push up Death Valley by frontal assault 
had failed again. In the words of Colonel 
Aycrs, "We were thrown right back on to 
the original line of departure." m 

Once again corps headquarters ordered 
the 27th Division to renew the attack into 
the valley next morning with the "main 
effort on the left." 102 General Ralph 
Smith, however, now decided that any 
further headlong rush up the valley would 
only result in increased casualties, and 
ordered his division to make the main effort 
on the right. 103 Beginning on the morning 
of the 25th, the 2d Battalion, 106th In- 
fantry, would take up positions along the 
entrance to the valley and contain the 
Japanese there while the 165th Infantry 

Love, Batt'c for Saipan, pp. 301-04, 
1 06th RCT Jnl, 24 Jun 44, Msg 308. 

ss Ibid., Msgs a6i, 285. 

99 Love, llattlp for Saipan, p. 304,, 

100 Ibid, 

101 Aycrs Testimony, Buckner Board Rpt, Ex- 
hibit UCC, p. 6. 

102 NTLF Opn Order 12-44, '8oo, 24 Jun 44. 

103 27th Inf Div FO 48, 141800 Jun 44, 27th 
Iilf Div G-3 Jnl, Incl. 



and the other two battalions of the 106th 
would circle around the cast (right) flank 
and come out into the valley at its northern 
end to the rear of the Japanese positions. 
This, it was hoped, would put the Army 
division abreast of the two Marine di- 
visions, and the encircled Japanese in 
Death Valley could then be mopped up at 
leisure. 104 

There is no evidence that this discrep- 
ancy between corps and division orders 
was ever noted at the time. In any case, 
before he could put this plan into execu- 
tion, General Ralph Smith was relieved of 
his command, and it was left to his suc- 
cessor, General Jarman, to solve the 
problem of Death Valley. 1015 

a 4 June: Action on the Flanks 

On the right of the corps line, the 4th 
Marine Division was ordered to press east- 
ward across Kagman Peninsula and secure 
that area before reorienting its drive 
toward the north of the island. The divi- 
sion jumped off on schedule at 0800, the 
24th Marines on the right, the 23d on the 
left. The 24th Regiment moved out rapidly 
along the coast to Kagman Peninsula 
against "moderate" resistance. By the end 
of the day the 24th Marines had advanced 
about 1 ,200 yards. 

On its left the 23d Marines was initially 
held up by a pocket of resistance on the 
slopes of Hill 600 that marked the boun- 
dary line between the Marine division and 
the Army division. About noon the Marine 
regiment detoured the pocket and com- 
menced to swing around the arc toward 
Kagman Peninsula, pivoting on the 24th 

Marines on its right. As the 23d Regiment 
accelerated its swing, the gap between it 
and the 165th Infantry increased and 
toward late afternoon amounted to from 
800 to 1,000 yards. By 1630 Chacha vil- 
lage was overrun and the advance halted 
for the night since the gap on the left 
precluded any further progress. The 1st 
Battalion, 23d Marines, in division reserve, 
was ordered to occupy Hill 600 and assist 
its parent regiment in patrolling the gap. 
Opposition in the 4th Division's zone on 
the 24th was characterized as "moderate" 
and "light," although total casualties came 
to 380, including killed, wounded, and 
missing in action. 10 * 5 

To the west (left) of Death Valley the 
2d Marine Division was drawn up abreast, 
from right to left, 8th Marines (with the 
1 st Battalion, 29th Marines, attached), 6th 
Marines, and 2d Marines. On the right 
flank the 1st Battalion, 29th Marines, was 
struggling along the cliff overlooking Death 
Valley. The ground was "a tangle of tree 
ferns and aerial tree-roots overgrown with 
a matting of vines. This formation led up 
to a ridge, access to which required an 
almost vertical climb." 10T By nightfall 
the battalion had succeeded in climbing 
the ridge, which connected with and was 
within machine gun range of the summit 
of Mount Tapotchau. The battalion's ad- 
vance for the day was about 800 yards. 108 

During this movement of the 1st Bat- 
talion, 29th Marines, Company F of the 
106th Infantry was virtually an integral 

104 Ralph Smith, Notes, Buckner Board Rpt, Ex- 
hibit M, p. 1 5. 

100 See belowlp. 203. 1 

1011 4th Marine Div Opns Rpt Saipan, pp. 24- 
35; NTLF Rpt Marianas, Phase I, Incl C, C-i 
Rpt, App, 1, Casualty Rpt; Hoffman, Saipan, pp. 

107 ad Marine Div SAR, Phase I, Forager, ScC. 
VI, p. 12. 

10S Ibid., p. 13. 



part of the Marine battalion and had been 
ordered by its own battalion headquarters 
to gear its movements to those of the 
marines. ,0B About 1230, F Company came 
to a halt when the marines on its left en- 
gaged in a fire fight. The company's right 
flank rested along the edge of the cliff over- 
looking Death Valley just opposite the 
north rim of Hell's Pocket. The terrain tea 
the right consisted of a down-sloping nose 
of ground that broke off abruptly in the 
cliffs that scaled down the valley. Capt. 
Roderick V. Le Mieux, the company com- 
mander, ordered a patrol to probe down 
this nose and investigate the source of en- 
emy rifle fire coming from that direction. 
The patrol soon flushed a covey of Japa- 
nese and a full-scale fire fight developed. 
The patrol leader, finding himself outnum- 
bered, returned for more men. With the 
larger group he again went down the nose 
and succeeded in driving the enemy back 
down the hill, 110 

Meanwhile, Captain Le Mieux had 
moved ahead a short distance and observed 
a Japanese artillery piece in a cave in the 
side of the cliff about 500 yards to his 
front. The Japanese were playing a hide 
and seek game with the piece, running it 
out of the cliff to fire at the line of trees 
along the south edge of the valley below, 
then quickly dragging it bark under cover. 
Le Mieux observed about a hundred of 
the enemy in the vicinity of this position as 
well as a large ammunition dump. He or- 
dered up heavy weapons from Company H, 
which succeeded in catching the gun out 
of the cave and destroying it. In addition 
they blew up the ammunition dump, killing 

about thirty Japanese. Shortly thereafter 
an enemy patrol was discovered advancing 
toward the heavy weapons outpost. Rifle 
fire and grenades quickly dispersed the 
Japanese, but not before Captain Le 
Mieux received a serious fragmentation 
wound and had to be replaced by 1st Lt. 
Herbert N. Slate, the company executive 
officer. A few minutes later the company 
was ordered to pull back to the left and 
to tighten up the lines before digging in 
for the night. 

In the center, the 8th Marines advanced 
with no particular difficulty, maintaining 
contact with units on both of its flanks. 
The battalion on the regimental left, on 
the other hand, after advancing a short 
distance, encountered heavy enemy resist- 
ance in an area honeycombed with caves 
and irregular coral limestone formations 
covered with trees and undergrowth. This 
was the same pocket that had retarded the 
battalion on the previous day. Contact 
with the 6th Marines on the left was tem- 
porarily broken. Shortly after noon the 
pocket was cleaned out and contact was 
restored with the regiment on the left. The 
advance continued until late afternoon and 
by the time the 8th Marines dug in for the 
night it had registered a day's gain of about 
700 yards. 111 

The advance of the 6th Marines in the 
division center was not so rapid, especially 
on the right flank where it faced cliffs and 
thickly wooded ravines and encountered 
strong enemy positions in natural cave 
formations. By evening the regiment had 
progressed from about 900 yards on the 
left to 500 to 600 yards on the right. The 
regimental lines had become too extended 

,0!> ad Tin 106th RCT Jul, 24 Jun 44, Msg gfi. 
1111 This account of F Company's activities is de- 
rived from Love, Battle for Saipan, pages 306-08. 

111 2d Marine Div SAR, Phase I, Forager, Sec. 
VI, p. ig; Hoffman, Saipan, p. 143. 




In Hellas Pocket Area 

for safety, and an additional company was 
committed to the line. 112 

On the division left the 2d Marines 
jumped off at 0800. On its right, progress 
was slow against heavy fire from a hill, just 
southeast of Garapan, that the marines did 
not occupy until 1500. Shortly thereafter 
the Japanese counterattacked the hill, 
which from their side (north) was virtually 
a cliff. Firing with muzzles depressed 
against the enemy below, the marines easily 
repulsed the attack and dug in for the 
night along the ridge overlooking "Radio 
Road," which ran at right angles to the 
line of advance. Meanwhile, the battalion 
on the left had quickly advanced 500 yards 

11 - 2d Marine Div SAR, Phase I, Forager, Sec. 
VI, p. .3. 

along the beach into the southern outskirts 
of Garapan itself. Around 1625, as these 
men were preparing their defenses along 
the flatlands bordering the sea, seven en- 
emy tanks unaccompanied by infantrymen 
suddenly moved out from Garapan against 
them. Marine tanks and 75-mm. half- 
tracks were rushed in and quickly broke 
up the attack, destroying six of the Japa- 
nese vehicles and routing the seventh. The 
2d Marines had now reached the 0-6 line 
and would again have to hold up until 
the units on its right came abreast. Cas- 
ualties for the entire 2d Division on the 
24th amounted to 31 killed, 165 wounded, 
and one missing in action. 113 

1 1S Ibid.; Hoffman, Saiparij pp. 141-42. 



Garapdn Pier 


Hill occupied by 
2d Marines 

Inland of Garapan Harbor from hill position southeast of the town, 24 June. 

After two days of fighting on a three 
divisional front, the attack of Holland 
Smith's corps against the center of the Jap- 
anese main line of resistance had stalled. 
On the right, the 4th Marine Division had 
overrun most of Kagman Peninsula, an 
area that presented no particular terrain 
problems but that still contained plenty of 
live Japanese, judging from the casualties 
suffered there by the marines. On the left, 
the 2d Marine Division had fought its way 
into the outskirts of Garapan and up the 
craggy approaches to Mount Tapotchau, 
although it would take another day to 
reach the summit of the mountain. In the 
corps center, the 165th Infantry on the 

27th Division's right had captured Hill 
Love, but had made no further advance 
along the hill system called Purple Heart 
Ridge, Progress of the 106th Infantry into 
Death Valley had been negligible. Thus, in 
the center — the 27th Division zone — the 
corps line bent back as much as 1,500 
yards. Total reported casualties in the 4th 
Marine Division for the two days were 812; 
for the 2d Marine Division, 333; for the 
27 th Infantry Division, 277. 114 

111 27th Infantry Division casualty figures are 
from 27th Inf Div G-i Periodic Rpts; 2d Marine 
Division casualty figures are from 2d Marine Div 
SAR Phase I, Forager, Sec. VI; 4th Marine Di- 
vision casualty figures are from NTLF Rpt Mari- 
anas, Phase I, Incl C, G-i Rpt. 


Smith Versus Smith 

Relief of Major General 
Ralph C. Smith 

By 34 June General Holland Smith had 
made up his mind that the "all-round poor 
performance" of the 27th Division could 
only be remedied by a drastic shake-up in 
its command structure. 1 Accordingly, he 
decided to ask for the relief of General 
Ralph Smith. 

He first visited Admiral Turner, who 
agreed with him, and together the two of- 
ficers boarded the flagship Indianapolis to 
consult with Admiral Spruance. As a re- 
sult of this discussion, Admiral Spruance 
"authorized and directed" that General 
Ralph Smith be relieved by General Jar- 
man, the island commander. It was under- 
stood that Jarman would take over only 
until such time as another general officer 
could be dispatched from Hawaii to com- 
mand the division. In Spruance's words, 
"No other action appeared adequate to 
accomplish the purpose," 2 

The bill of particulars presented by Gen- 
eral Holland Smith against General 
Ralph Smith broke down into two general 
charges : ( 1 ) that on two separate oc- 
casions the Army commander had issued 

1 The words arc General Holland Smith's. See 
Smith, Coral and Brass, p. 171. 

2 Ltr, Comdr Fifth Fleet to CINCPAC- 
CINGPOA, 29 Jun 44, Ser. 00025, Buckner Board 
Rpt, Exhibit C. 

orders to units not under his command and 
had contravened orders of the corps com- 
mander; and (2) that on the morning of 
23 June the 27th Division had been late 
in launching its attack and had thereafter 
retarded the progress of the Marine divi- 
sions on the flanks."' 1 

On the first point, the corps commander 
charged that the "27th Infantry Division 
Field Order No. 43 d contravened the 
NTLF Operation Order Number 9-44 by 
ordering the 105th Infantry to hold its 
present positions, although the 105th In- 
fantry had been removed from the tactical 
control of the Division Commander," and 
that the 27th Division "Field Order No. 46 
again contravened the NTLF order by is- 
suing tactical orders to the 2d Battalion, 
iof)th Infantry to continue operations to 
mop up enemy resistance in NAFUTAN 
POINT Area," although that battalion "by 
NTLF Operation Order No. 10-44 had 
been removed from the tactical control of 
the 27th Infantry Division." 5 

On the second point, it was alleged that 
on the morning of 23 June, the "27th In- 
fantry Division was from 77 minutes to two 
hours late in launching its attack, although 
the major elements of this division did not 

* Ltr, GTF 36 to CTF 51, 24 Jun 44, Ser. 
00055-3, Buckner Board Rpt, Exhibit D. 

4 He meant Field Order Number 45-A. 

5 Ltr, CTF 56 to CTF 51, 24 Jun 44, Ser. 
00055-3, Buckner Board Rpt, Exhibit D. 



have to move more than about three miles 
to execute the order." 6 In a report to Ad- 
miral Turner written three days later, Gen- 
eral Holland Smith revised this figure 
downward to "55 minutes to two hours" 
and added that the "lack of coordination 
in the attack" resulting from the 27th Di- 
vision's late arrival and "the slow advance 
of the Division against small arms and 
mortar fire uncovered the flanks of the 4th 
and 2d Marine Divisions to such extent that 
it was necessary to slow down and eventu- 
ally halt these units and thereby retard 
otherwise favorable offensive operations 
which were in progress." 7 

Interservice Controversy 

It is doubtful whether the relief of Gen- 
eral Ralph Smith brought about any 
marked change one way or the other in the 
"aggressiveness" of the 27th Division about 
which General Holland Smith was so con- 
cerned. 8 There is no doubt, however, that 
it precipitated an interservice controversy 
of alarming proportions — a controversy 
that seriously jeopardized harmonious re- 
lations at all levels among the Army and 
the Navy and the Marine Corps in the 

The first signs of strain appeared nat- 
urally enough on Saipan itself, where 
soldiers and marines still had to fight 
shoulder to shoulder for more than two 
weeks to secure the island. Army officers 
were quick to resent the slur on their 
service implied by the relief of General 
Ralph Smith, and by the end of the battle 

« Ibid. 

7 Ltr, CG NTLF to CTF 51, 27 Jun 44, Ser. 
00063-3, Wll h Incl A, Buckner Board Rpt, Ex- 
hibit V 

s See below, Chs. XI, XIII, 

relationships between top Army officers 
and Holland Smith's staff had reached the 
breaking point. Various Army officers who 
had contact of one sort or another with 
that staff reported that the Marine officers 
at headquarters made little effort to dis- 
guise their feeling that the 27th Division 
was an inferior organization. In the opinion 
of one of the Army officers, "the Com- 
manding General and Staff of the NTLF 
held the units of the 27th Division in little 
esteem, actually a position bordering on 

The reaction on the part of the ranking 
Army officers present on Saipan was a de- 
termination never to serve under General 
Holland Smith again if they could help it. 
Gen. Ralph Smith urged Lt. Gen, Robert 
G Richardson that "no Army combat 
troops should ever again be permitted to 
serve under the command of Marine Lieu- 
tenant General Holland M. Smith." I0 
General Kcrnan, who commanded the 27th 
Division Artillery, agreed. 11 Maj. Gen. 
George W. Grincr, who took over com- 
mand of the Army division on 26 June, 
quarreled so bitterly with the corps com- 
mander that he came away from Saipan 
with the "firm conviction that he [Hol- 
land Smith] is so prejudiced against the 

n Ltr, Maj Stephen J. McCormick, SC, to CG 
Army Garrison Force, Saipan, a Sep 44, sub: Re- 
flections on 37th Inf Div, in Hq AFMIDPAC, 
Correspondence Incident to Relief of Maj Gen 
Ralph C, Smith, 27th Division (hereafter cited 
as AFMIDPAC Corresp, Ralph Smith) bound 
photostatic copy in OCMH. See also Brig Gen 
Edgar 11. Colkiday, USA, Memo for Island Comdr 
Saipan, f> Sep 44; Lt Col P. B. Stiness, GSC, 
Memo for CG Army Garrison Force, APO 244, 4 
Sep 44, both in corresp cited above. 

10 Maj Gen Ralph C. Smith, Recommendation 
to CG USAFICPA, 28 Aug 44, in AFMIDPAC 
Corresp, Ralph Smith. 

11 Ltr, Kernan to Richardson, 16 Aug 44, in 
AFMIDPAC Corresp, Ralph Smith. 



Army that no Army Division serving under 
his command alongside of Marine Divisions 
can expect that their deeds will receive fair 
and honest evaluation." 12 

When, less than a week after the con- 
clusion of organized hostilities on Saipan, 
the island was visited by General Richard- 
son, the commanding general of all Army 
forces in the Pacific Ocean Areas, the dis- 
pute waxed even hotter. While on the is- 
land, Richardson reviewed the Army troops 
and presented decorations — all without the 
previous knowledge or consent of Holland 
Smith. 11 ' The corps commander was quick 
to resent these actions, which he considered 
to be a breach of military etiquette and an 
unwarranted infringement on his own au- 
thority. On his part, General Richardson 
is reported to have said angrily to the Ma- 
rine general, "I want you to know you can 
not push the Army around the way you 
have been doing." 14 At this juncture Ad- 
mirals Spruance and Turner jumped into 
the fight and complained strongly to 
Admiral Nimitz of the irregularity of 
Richardson's actions on Saipan, and es- 
pecially his berating of Holland Smith. lfi 

General Richardson's visit to Saipan was 
in fact incident to a more general inquiry 
into the relief of Ralph Smith, which 
Richardson had called at his headquarters 
back on Oahu. On 4 July, five days before 
the conclusion of the battle for Saipan, 

12 Ltr, Griner to CG USAFPOA, n Oct 44, in 
AFMIDPAC Corresp, Ralph Smith. 

L:f One account of this visit is given by General 
Holland Smith in Coral and Brass, pages I 76-78, 

14 CG Fleet Marine Force, Pacific, Rpt of Visit 
of Lt Gen Robert G. Richardson, Jr., 18 Jul 44, 
filed in RS Hist Br G-3, Hq USMC. 

15 Ltr, Comdr V Phib Force to CINCPAC, t6 
Jul 44, with COMCENPAC 1st Ind, 18 Jul 44, 
sub: Rpt of Unwarranted Assumption of Gomd 
Authority by Gen Richardson, filed in RS Hist 
Br G-3, Hq USMC. 

Maj. Gen. Sanderford Jarman, talk- 
ing with Admiral Spruance, June 1944. 

Richardson had appointed a board of in- 
quiry to examine the facts involved. 16 
The board was headed by Lt. Gen. Simon 
B. Buckncr, Jr., and consisted, in addition 
to the chairman, of four Army officers, 
Maj. Gen. John R. Hodge, Brig. Gen. 
Henry B. Holmes, Jr., Brig. Gen. Roy E. 
Blount, and Lt. Col. Charles A. Selby. It 
convened first on 7 July and continued un- 
til the 26th, hearing the testimony of Army 
officers and examining those official reports 
from Army files that were available to it. 17 
After examining all the available evi- 
dence — which was admitted to be limited 

'" Rad, Richardson to WDCSA, 4 Jul 44, in 
AFMIDPAC Corresp, Ralph Smith. 

' 'Buckner Board Rpt, Ralph Smith Personal 




Lt. Gen. Holland M. Smith 

because only personnel and records of the 
U.S. Army Forces, Central Pacific Area, 
could be examined — the "Buckner Board" 
arrived at four conclusions : 

i. that General Holland Smith had full 
authority to relieve General Ralph Smith; 

2. that the orders effecting the change of 
command were properly issued; 

3. that General Holland Smith "was not 
fully informed regarding conditions in the 
zone of the 27th Infantry Division," when 
he asked for the relief of General Ralph 
Smith; and 

4. that the relief of General Ralph Smith 
"was not justified by the facts." 18 

In reaching these conclusions, the Buck- 
ner Board reasoned that the situation facing 

ls Ibid., p. 10. 

the 27th Division at the entrance to Death 
Valley was far more serious than General 
Holland Smith had imagined. "The bulk 
of the 27th Division," the board reported, 
"was opposed by the enemy's main defen- 
sive position on a difficult piece of terrain, 
naturally adapted to defense, artificially 
strengthened, well manned and heavily 
covered by fire." General Holland Smith, 
it concluded, "was not aware of the 
strength of this position and expected the 
27 th Division to overrun it rapidly .... 
The delay incident to this situation was 
mistaken by Lt. Gen. Holland M. Smith 
as an indication that the 27th Division was 
lacking in aggressiveness and that its com- 
mander was inefficient . . . ," 10 Further- 
more, the board argued, there was no 
evidence that General Ralph Smith at- 
tempted to "contravene" orders during the 
clean-up on Nafutan Point. 

These findings, coming as they did from 
an all- Army board of inquiry by no means 
ended the controversy. Holland Smith 
wrote to Admiral Nimitz to the effect that 
the Buckner Board's conclusions were un- 
warranted, and added, "I was and am con- 
vinced that the 27 th Division was not 
accomplishing even the combat results to 
be expected from an organization which 
had had adequate opportunity for train- 
ing." 20 Admiral Turner, resenting the 
board's implied criticism that he had been 
overzealous in "pressing Lt. Gen. Holland 
M. Smith ... to expedite the conquest 
of Saipan so as to free the fleet for another 
operation," 2l also demurred from the find- 
ings of the board. He at no time had 
brought pressure to bear on Holland 

1H Ibid., -p. 3. 

-" CG Fleet Marine Force, Pacific, to ClNCPAC 
29 Aug 44, Proceedings of the Buckner Board, RS 
Hist Br G^s, Hq USMC 

?1 Buckner Board Rpt, p. 3. 



Smith, he asserted, and he was confident 
that no part of the Marine general's action 
against Ralph Smith "was based on 
either personal or service prejudice or 

When the detailed report of the proceed- 
ings of the Buckner Board reached Wash- 
ington, General Marshall's chief advisers 
tended to take a "plague on both your 
houses" attitude, Maj. Gen. Thomas T. 
Handy, Assistant Chief of Staff, advised 
Marshall that Holland Smith had some 
cause for complaining of the 27th Divi- 
sion's lack of aggressiveness in the attack 
into Death Valley; that "Holland Smith's 
fitness for this command is open to ques- 
tion" because of his deep-seated prejudice 
against the Army; and that "bad blood 
had developed between the Marines and 
the Army on Saipan" to such a degree that 
it endangered future operations in the the- 
ater, "In my opinion," he concluded, "it 
would be desireable that both Smiths be 
ordered out of the Pacific Ocean Area. 
While I do not believe we should make 
definite recommendation to the Navy for 
the relief of Holland Smith, I think that 
positive action should be taken to get Ralph 
Smith out of the area. His presence un- 
doubtedly tends to aggravate a bad situa- 
tion between the Services." 23 

Lt. Gen. Joseph T, McNarney, Deputy 
Chief of Staff, was of much the same mind 

"Comdr V Phib Force to CINCPAC, 18 Aug 
44, RS Hist Br G-3, Hq USMC. 

2 * Memo, Handy for CofS, 16 Aug 44, atchd to 
Buckner Board Rpt. This recommendation was 
acted upon favorably. Ralph Smith was relieved 
of his command of the 98th Infantry Division, 
which was on garrison duty in the Hawaiian Is- 
lands. He was later transferred to the European 
Theater of Operations. Holland Smith, while re- 
lieved of his command of V Amphibious Corps, 
was elevated to the command of the newly organ- 
ized Fleet Marine Force, Pacific. 

Lt. Gen. Robert C. Richardson, Jr. 

as General Handy. After examining the 
Buckner Board Report, he concluded that 
the staff work of Holland Smith's V Am- 
phibious Corps was below acceptable 
standards; that there was reasonably good 
tactical direction on the part of Ralph 
Smith; and that Ralph Smith failed to ex- 
act the performance expected from a well- 
trained division, as evidenced by poor 
leadership on the part of some regimental 
and battalion commanders, undue hesi- 
tancy to bypass snipers "with a tendency to 
alibi because of lack of reserves to mop 
up," poor march discipline, and lack of 
reconnaisance. 2 * 

On 22 November General Marshall ex- 
pressed to Admiral King his deep concern 
over the fact that "relationships between 

2i Memo, McNarney for Handy, 19 Aug 44. 
atchd to Buckner Board Rpt. 



the Marines and the Army forces on Sai- 
pan had deteriorated beyond mere healthy 
rivalry. To avert future controversies of the 
same sort, General Marshall suggested that 
he and Admiral King send identical tele- 
grams to Richardson and Nimitz adjuring 
them "to take suitable steps to promptly 
eradicate any tendency toward . . . dis- 
harmony among the components of our 
forces." Marshall also suggested that both 
commanders should conduct an immediate 
investigation into the Saipan affair with 
an eye to preventing the recurrence of any 
such imbroglio in the future. yr ' To this Ad- 
miral King replied that in his mind the 
findings of the Buckner Board were uni- 
lateral and suspect, and that the record 
improperly included intemperate attacks 
on the personal character and professional 
competence of General Holland Smith. He 
could not concur in any further investiga- 
tions in which General Richardson was to 
be a party because he felt that that officer 
had already done enough damage by his 
"investigational activities during his visit to 
Saipan" and by convening the Buckner 
Board. 20 There the matter was dropped as 
far as official action was concerned. 

The American public, however, was not 
to be permitted any early respite from the 
heated journalistic dispute that followed 
Ralph Smith's relief. First among the news- 
papers to air the matter was the Hearst 
press. Various affiliates of that syndicate 
pointed editorially to two lessons from the 
battle for Saipan. First, it was claimed that 

' ij Memo, Marshall for King, a a Nov 44, 
WDCSA 000.7 and draft of proposed telegram to 
Richardson from Marshall (a similar telegram to 
be sent by King to Nimitz), 

' 2fi Memo, COMINCH for Marshall, 23 Nov 44, 
sub: Article in Time Magazine, 12 Sep 44, 
WDCSA 000.7. 

Marine Corps casualties were excessive, es- 
pecially in contrast to those in Mac-Arthur's 
theater. Second, divided command was a 
mistake. The Hearst papers' conclusion 
was that "the supreme command in the 
Pacific should, of course, be logically and 
efficiently entrusted to General Douglas 
MacArthur." ' 2l 

Another powerful syndicate, the Henry 
Luce publications, took the other side. 
Time and Life magazines both carried ar- 
ticles favoring Holland Smith's side of the 
controversy, the former concluding, "when 
field commanders hesitate to remove sub- 
ordinates for fear of interscrvice contention, 
battles and lives will be needlessly lost." 2S 

More than four years after the event, the 
issue was reopened publicly when General 
Holland Smith published part of his war- 
time memoirs in the Saturday Evening 
Post. 29 He was answered by Capt. Ed- 
mund G. Love, the official historian of the 
27 th Infantry Division, in a rebuttal that 
was printed in part in the Saturday Eve- 
ning Post, and in full in the Infantry 
Journal* The capstone of this particular 
literary controversy was inserted when 
General Holland Smith published his mem- 
oirs in book form in 1949, and Captain 
Love in the same year came out with the 

11 San Francisco Examiner, July 6, 1944, p- 5- 
Sec also New York Journal American, July 17—18, 

28 Time, September 18, 1944, pp. 126-27. See 
also Life, August 28, 1944, pp. 33-42- 

an General Holland M. Smith and Percy Finch, 
"Howlin' Mad's Own Story," a book condensation 
in three parts, Saturday Evening Post, November 
6, 13, 20, 1948. 

so Edmund G. Love, Official Historian of the 
27th Division, "The Army Says Holland Smith Is 
Wrong," Saturday Evening Post, November 13, 
194H, pp. 33, f>5; Edmund G. Love, "Smith versus 
Smith," Infantry journal, LXIII (November, 
'948), 3-13. 



official history of 
World War II. 31 

the 27 th Division in 


To resolve the controversy of Smith 
versus Smith conclusively and to the satis- 
faction of all is probably impossible. But a 
dispassionate re-examination of the salient 
facts of the case as presented in the fore- 
going chapters may serve at least to clarify 
the issue and to point to some satisfactory 

The first charge against Ralph Smith 
dealt with his alleged usurpation of author- 
ity and contravention of orders in handling 
the troops of the 27 th Division that were 
left to finish the capture of Nafutan 
Point. In order to examine this charge it 
will be necessary first to recapitulate some 
of the events that took place on 21 and 22 

It will be remembered that on the morn- 
ing of 21 June Holland Smith issued 
Operations Order Number 9-44, which di- 
rected that the bulk of the 27th Infantry 
Division be removed from the front lines 
on Nafutan peninsula and be assembled 
northwest of Aslito field in corps reserve. 
In Paragraph 3(d) of this operations or- 
der, one infantry battalion (undesignated) 
of the division was ordered to remain on 
Nafutan peninsula, where it would "mop 
up remaining enemy detachments, main- 
tain anti-sniper patrols . . . and protect 
installations within its zone of action 
with particular attention to ASLITO Air- 
field." ;,a 

After an afternoon in which his troops 
made little progress on Nafutan, Ralph 

31 Smith, Coral and Brass; Love, The sylh In- 
fantry Division in World War II. 

3 - NTLF Opn Order 9-44, si Jun 44; 27th Inf 
Div G-3 Jnl, 2 1 Jun 44, Msg 34. 

Mat. Gen, Ralph C, Smith 

Smith called Holland Smith and persuaded 
him that at least two battalions would be 
needed to mop up the enemy in that area. 
Accordingly, the corps commander modi- 
fied his initial order in a mail brief that 
arrived at 27th Division headquarters at 
0830 on 22 June. This message read, "1 
RCT will continue mission in Garrison 
Area [Nafutan] of cleaning up remain- 
ing resistance and patrolling area . , . ," 3rf 
Like the initial order, this mail brief did 
not specifically designate the unit intended 
for the mission, although it was understood 
from previous conversations that the 105th 
Infantry would be given the job. 

At 2000, 2i June, after his conversation 
with General Holland Smith but before re- 
ceiving the mail brief modifying Operations 
Order Number 9-44, General Ralph Smith 

3S 27th Inf Div G-3 Jnl, 22 Jun 44, Msg 14. 



issued his Field Order Number 45-A. This 
order, insofar as it applied to the 105th 
Infantry, read: 

RCT 105 will hold present front line fac- 
ing NAFUTAN POINT, with two Battalions 
on the line and one Battalion in Regimental 
Reserve. It will relieve elements of RCT 165 
now on the present front line by 0630 33 
June. The Battalion in reserve will not be 
committed to action without authority from 
the Division Commander. Reorganization of 
the present front line to be effected not later 
than 1 100 22 June and offensive operations 
against the enemy continued. Reserve Bat- 
talion will maintain anti-sniper patrols in vi- 
cinity of ASLITO AIRFIELD. 34 

In asking for the relief of Ralph Smith, 
Holland Smith claimed that in issuing this 
field order, the 27th Division commander 
had committed two offenses simultane- 
ously. He had usurped authority of his 
immediate superior by issuing formal or- 
ders to a unit no longer under his control, 
and he had contravened his superior's 
orders by instructing that unit to "hold" 
rather than to fight offensively. Holland 
Smith argued that his corps Operation 
Order Number 9-44, as modified by the 
mail brief, placed the entire 27th Division 
in reserve status and removed the 105th 
Infantry from tactical control of the 27th 
Infantry Division, Hence, Ralph Smith 
had no right at all to issue orders to 
the 105th. Furthermore, Holland Smith 
claimed, his own order directed the 105th 
Infantry "to conduct offensive operations 
to mop up enemy units in the NAFUTAN 
POINT area." 35 Ralph Smith's Field Or- 
der Number 45-A, on the other hand, in- 

u * 37th Inf Div G-3 Rpt, Battle for Saipan, 
Field Directive 4.5, Able, 2000, 21 Jun 44. Italics 
are the author's. Where the order used abbrevia- 
tions, the words have, been spelled out. 

35 Ltr, CTF 56 to CTF 51, 24 Jun 44, Buckner 
Board Rpt, App. D. 

structed the 105th Infantry "to hold its 
present positions" rather than to conduct 
offensive operations. This, according to 
Holland Smith, was a clear contravention 
of orders. 

Both Army and Marine Corps regula- 
tions concerning the composition of combat 
orders tend to support Holland Smith's 
argument on the question of where control 
of the 105th Infantry lay on the night of 
21 June. Furthermore, they account in 
part for his own conviction that tactical 
control over the 105th had been clearly 
removed from the 27 th Division and had 
been placed under his own headquarters 
by his Field Order Number 9-44. These 
regulations state that Paragraph 3 of a 
field order "assigns definite missions to each 
of the several elements of the [issuing] 
command charged with execution of the 
tactical details for carrying out the decision 
of the commander or the assigned mis- 
sion." 36 Since the "one Infantry Battal- 
ion, 27th Infantry Division (to be desig- 
nated)," was assigned a specific mission in 
Paragraph 3(d) of Holland Smith's Field 
Order Number 9-44 and since the entire 
105th Infantry was shortly thereafter 
substituted for this one battalion, it seemed 
clear to members of Holland Smith's staff 
that the unit would execute its mop-up 
task as an immediate subordinate of Hol- 
land Smith's headquarters. 37 

General Ralph Smith, on the other hand, 
was just as clear in his mind that the unit 
left on Nafutan was still under his own 
command. Speaking of his telephone con- 
versation with General Holland Smith, he 
later recollected, "Nothing was mentioned 

.16 \y ar rjept FM iOi-5, 19 Aug 40, p. 43. This 
manual was in effect in 1944 and governed both 
Army and Marine Corps procedures. 

ST Hoffman, Saipan, p. iaqn. 



in his conversation about having the regi- 
ment [loSthJ operate under NTLF con- 
trol," 3S He continued that, in his opinion, 
his Field Order Number 45-A was neither 
a usurpation nor a contravention of orders. 
No written confirmation of the mission to 
be assigned to the 105th Infantry arrived 
until 0830, 22 June, much too late to have 
permitted issuing any instructions for that 
day's operation. The 105th Infantry was to 
take over with two battalions a front line 
covered the previous day by four bat- 
talions. "It seemed elementary military 
common-sense to have these two battalions 
first take over the front from the units 
being relieved." Hence, in the absence of 
any further orders from higher headquar- 
ters, at 2000 on the night of the 21st 
Ralph Smith had ordered the 105th to 
"hold present front line," relieve elements 
of the 165th Infantry, and jump off not 
later than 1 1 00 the following morning. 
"The 105th Infantry was thus directed to 
resume offensive operations as soon as the 
lines were adjusted, thus to carry out the 
plan recommended by me and approved by 
General Holland Smith." 30 

Two facts stand out in support of Gen- 
eral Ralph Smith's contention. In the first 
place, Corps Order Number 9-44 did not 
specifically and expressly detach the 105th 
Infantry from the 27th Division and attach 
it to corps. Secondly, neither Corps Or- 
der Number 9-44 nor the subsequent 
mail brief mentioned the regiment by 
name, nor is there any record that either 
was sent to the command post of that reg- 
iment. Presumably, had General Ralph 
Smith not issued his Field Order Number 

45-A, the 105th Infantry would have been 
without orders for 22 June. 40 

On the afternoon of 22 June, General 
Holland Smith decided that a single bat- 
talion would be sufficient to clean up Na- 
futan Point. His chief of staff, General 
Erskine, personally communicated this de- 
cision to General Ralph Smith, That even- 
ing, the 27th Division commander drew up 
his Field Order Number 46, which he is- 
sued at 2100. In part, the order read: "2d 
Battalion, 1 05th Infantry ( 1 Platoon Light 
Tanks attached) [will] continue opera- 
tions to mop-up remaining enemy detach- 
ments in NAFUTAN POINT area. On 
completion of this mission, [it will] revert 
to Corps control as Corps Reserve." 41 

Just one hour later, Holland Smith is- 
sued his Operations Order Number 10-44, 
which was not received at 27th Division 
headquarters until 2330. 43 This order read 
in part: "2d Battalion 105th Infantry 
(with one light tank platoon attached) 
[will] continue operations at daylight to 
mop up remaining enemy detachments in 
NAFUTAN POINT area. Upon comple- 
tion this mission [it will] revert to Corps 
control as Corps reserve." 

In requesting the relief of Ralph Smith, 
Holland Smith alleged that the Army gen- 
eral's Field Order Number 46 contravened 
Corps Order Number 10-44 "by issuing 
tactical orders to the 2d Battalion, 105th 
Infantry, to continue operations to mop up 
enemy resistance in NAFUTAN POINT 
area. The 2d Battalion, 105th Infantry, 
by NT and LF Order No. 1 0-44, had been 

, '' 8 Ralph Smith, Notes, Saipan, Buckner Board 
Rpt, Exhibit M, p. 7. 
3B Ibid. 

*° The 105th RCT Journal for ii and 11 Juiie 
1944 contains no record of the receipt of Corps 
Order Number 9-44. 

41 27th Inf Div FO 46, a 100, 22 Jun 44, in 27th 
Inf Div G-j Rpt, Battle for Saipan. 

42 NTLF Opn Order 10-44, 2200, 22 Jun 44; 
27th Inf Div G-3 Jnl, 22 Jun 44, Entry 78. 



removed from the tactical control of the 
27th Infantry Division." 43 

Actually, of course, the only difference 
between Ralph Smith's Field Order 
Number 46 and Holland Smith's Order 
Number 10-44 in respect to the 2d Bat- 
talion, 105th Infantry, is that the latter 
included the words "at daylight" and the 
former omitted them. Otherwise, they are 
identical in all essential points. Later, Ralph 
Smith testified that in his conversations 
with General Holland Smith up to date 
no mention had been made of any question 
of control of the 105th Infantry nor had 
he been given any indication that that unit 
was no longer under direct control of the 
27th Division. 44 His belief that the 2d Bat- 
talion, 105th, was still under his tactical 
control was reinforced by the wording of 
Corps Order Number 10-44 itself. The 
fact that the order stipulated that "upon 
completion this mission" the battalion was 
to "revert to Corps control as Corps re- 
serve" would seem to indicate strongly that 
until its mission was completed, the unit 
was not under corps control but still under 
the division. 

The fact is that the orders from Holland 
Smith's headquarters were never clear as 
to where command authority over the 
troops on Nafutan Point did lie. Ralph 
Smith had to issue some orders, or none 
would have reached the front-line troops 
in time. There was no important difference 
between the commands that he issued and 
those that later came down from corps 
headquarters. There is no indication that 
any "contravention" of orders was in- 
tended or effected. At best, this charge ap- 

43 CG Expeditionary Troops to CO TF 51, 24 
Jun 44, Buckncr Board Rpt, Annex D, p. 2. 

14 Ralph Smith, Notes, p. 11, Buckner Board 
Rpt, Exhibit M. 

pears to have been a rather flimsy legal 
peg upon which to hang a justification for 
Ralph Smith's relief. 

The second charge was more serious. It 
concerned the tardiness of the 27th Divi- 
sion in jumping off into Death Valley on 
the morning of 23 June, the alleged poor 
co-ordination of the division in the attack, 
and its slow advance against "small arms 
and mortar fire," which slowed down the 
whole corps attack. Connected with this 
charge was Holland Smith's opinion, as 
later expressed, that the Army division was 
guilty of "all-round poor performance." 45 
Here was undoubtedly the core of Holland 
Smith's complaint against the 27 th Infan- 
try Division and its commander, and it is 
on these allegations that the case between 
him and Ralph Smith must be decided. 

The details of the fighting at the en- 
trance to Death Valley on 23 and 24 June 
have already been presented. 46 Out of 
this complex of events, several conclusions 
emerge. On the one hand, it appears clear 
that Holland Smith and his staff underes- 
timated both the formidability of the ter- 
rain and extent of enemy opposition that 
faced the 27th Division in Death Valley 
on the days in question. 

The terrain facing the 27th Division was 
most difficult. Two parallel ridges on the 
division flanks dominated its zone of action, 
and flanking fire from well-concealed en- 
emy positions on the slopes interdicted the 
valley between the ridges. Before the divi- 
sion could accomplish its mission the enemy 
occupying these dominant terrain features 
had to be eliminated. 

45 CG NTLF to CTF 51, 27 Jun 44, Buckner 
Board Rpt, IncI B; Smith, Coral and Brass, p, 

4,1 See above, Ch. IX. 



The conditions obtaining in the left part 
of the division zone precluded the possibil- 
ity of maneuver, and an attack along the 
east slopes of Mount Tapotchau would 
have to be a frontal assault. Because of 
extremely rugged terrain, flanking enemy 
fire from Purple Heart Ridge, and the 
difficulty of co-ordination with the Marines 
on the left, any such frontal attack would 
necessarily be costly. 

In the right part of the division zone the 
terrain was less rugged, and, more impor- 
tant, there was a possibility of a flanking 
maneuver east of Purple Heart Ridge. This 
was clearly the more promising area for 
the main attack by the Army division. Yet 
even as late as the evening of 24 June after 
two days of heavy and generally fruitless 
fighting on the part of the 27th Division, 
corps headquarters still ordered the main 
effort to continue on the left. 47 

On the other hand, there is no doubt 
that the 106th Infantry Regiment of the 
27th Division was late in jumping off in 
the attack on the morning of 23 June — 
even though not so late as Holland Smith 

47 See above, p. 186. 

charged. On the 23d and again on the 
24th, the Army troops attacking Death 
Valley were slow and faltering in their ad- 
vance. According to the testimony of Gen- 
eral Jarman, who took over the division 
from Ralph Smith, the unit leaders of the 
106th Infantry were hesitant and appar- 
ently confused. Although the Army troops 
in Death Valley sustained fairly heavy 
casualties, the two Marine divisions on the 
flanks suffered greater ones. Yet the ma- 
rines made considerable advances while the 
165th Infantry registered only small gains 
— the 106th Infantry almost none at all. 
No matter what the extenuating circum- 
stances were and there were several — the 
conclusion seems inescapable that Holland 
Smith had good reason to be disappointed 
with the performance of the 27th Infantry 
Division on the two days in question. 
Whether the action he took to remedy the 
situation was a wise one, however, remains 
doubtful. Certainly the relief of Ralph 
Smith appears to have done nothing to 
speed the capture of Death Valley. Six 
more days of bitter fighting remained be- 
fore that object was to be achieved. 


/ r^v X 

/ / \ 




The Fight for Central Saipan— II 

With General Ralph Smith's relief ef- 
fected, General Jarman received orders to 
take over command of the 27th Division. 
He reported to division headquarters in the 
middle of the afternoon of 24 June and dis- 
covered that General Ralph Smith was at 
the front inspecting troops. Late that after- 
noon General Smith returned to the 
command post and remained there in con- 
sultation with General Jarman until about 
0100 the next morning. The plan agreed 
upon for 25 June was essentially the one 
General Smith had already devised. Rather 
than continue the frontal assault on Death 
Valley with all three battalions of the 106th 
Infantry, it was decided that one battalion 
would be left at the mouth of the valley to 
contain the Japanese while the other two 
would circle to the right (eastward), then 
turn northwest and establish contact with 
the ad Marine Division north of the Jap- 
anese positions that had held up the 27th 
Division's advance through the valley. 1 
Early on the morning of the 25th, while 
the two generals were still together, Gen- 
eral Smith received his orders to report not 
later than 0530 that day for air transporta- 
tion back to Pearl Harbor. 2 

1 Jarman, Memo for Record, 24 Jun 44, Buck- 
ncr Board Rpt, Exhibit J; Ralph Smith Notes, 
Buckner Board Rpt, Exhibit M, p. 16. 

3 CG NTLF to Ralph Smith, 25 Jun 44, Ser. 
2445, Buckner Board Rpt, Exhibit O. 

25 June 

General Jarman issued his orders for the 
25th soon after his long conversation with 
General Ralph Smith. He directed the 
165th Infantry to continue its advance and 
seize the O-5 line in its zone, including all 
of Purple Heart Ridge. Jump-off hour was 
to be 0730. One battalion was to mop up 
the Japanese on Hill Xray-Yoke and in the 
gulch to the south of it while the other 
two were to move up the ridge itself. Start- 
ing at 0600 the 1 st and 3d Battalions, 
106th Infantry, were to move by covered 
route in the rear of the 165th, then ad- 
vance to the northwest, cross the northern 
entrance of the valley, and establish con- 
tact with the 2d Marine Division, The 2d 
Battalion, 106th, was ordered to remain in 
its current position at the southern end of 
Death Valley, conduct mop-up operations, 
and "assist in containing and eliminating 
enemy positions" within the valley. Divi- 
sion artillery was to fire a fifteen-minute 
preparation. 3 XXIV Corps Artillery would 
be in general support and was ordered to 
place the mass of its fires in the zone of 
the 27th Division. 4 \Map 8)\ 

3 27th Inf Div FO 46, 242100 Jun 44, Incl to 
27th Inf Div G-3 Jnl. 

* NTLF Opn Order 1 2-44, 24 Jun 44. 



Attack on Purple Heart Ridge 

Early on the morning of the 25th, 
Colonel Kcllcy issued orders to his three 
battalions to move northeast up Purple 
Heart Ridge. The 2d Battalion, 165th, was 
to attack Hill Xray-Yoke frontally from its 
position at the foot of the hill and then 
move on to Hills Oboe and King beyond. 
The 3d Battalion was to move up on the 
right of the 2d Battalion so that the two 
could jump off together at 0730. The 1st 
Battalion was to mop up all enemy resis- 
tance behind the two assault battalions and 
follow them northward. 5 Had these orders 
been carried out and had the 1 st Battalion, 
106th, succeeded in tying in with the right 
flank of the 165th as planned, the assault 
on Purple Heart Ridge would have been 
conducted as a co-ordinated movement 
with three battalions abreast and one in 
support. As it turned out, the fighting 
degenerated into separate and unco- 
ordinated actions by each of the four 
battalions involved. 

The 1st Battalion, 165th, moved off on 
schedule from the base of Hill Love and 
after a short fire fight cleared the last ves- 
tige of the enemy from the area to the 
north. The battalion combed the area in 
its immediate zone without incident and 
then retired to its bivouac area of the two 
previous nights without taking any further 
part in the day's attack on Purple Heart 

The immediate task of the 2d Battalion, 
165th, was to take Hill Xray-Yoke, which 
Company K had unsuccessfully attacked 
the previous day. Company F had biv- 
ouacked the night before southwest of the 
gulch below the hill. Instead of attempting 

another frontal attack from that direction, 
Captain Leonard conducted a wide detour 
and came upon Hill Xray-Yoke from a 
southeasterly direction. Company G, af- 
ter investigating the gulch itself, moved 
directly up the face of the hill from the 
south. 7 Company F arrived at the eastern 
foot of Hill Xray-Yoke about 0930 s and 
commenced to move up the slope in column 
of platoons, led by 1st Lt. Ford Martin, 
Captain Leonard, meanwhile, had joined 
the battalion commander, Colonel Mc- 
Donough, who was conducting a separate 
reconnaissance for the purpose of locating 
a site to establish an observation post. 
McDonough was desperately trying to 
close the gap between his own troops and 
the Marine line to the northeast. Just two 
hundred yards to the cast he could observe 
a road so congested with American troops 
and vehicles that it reminded him of Times 
Square on New Year's Eve. "It was a ma- 
chine gunner's dream," he later recalled, 
"but not one shot was being fired at it 
from the ridge to my front and this ridge 
showed no signs of life." B 

As McDonough's party reached the top 
of the hill, just before the main body of 
the company was approaching it, a ma- 
chine gun opened up and felled the entire 
group. 10 McDonough was wounded, as 
were Captain Leonard, Lieutenant Martin, 
and five enlisted men of F Company. 1st 
Lt. Henry W, Morrow, also of F Company, 
was killed. 11 Captain Leonard later died 

5 165th RCT Jnl, 25 Jun 44, Msgs 1, 2. 
K Love, Battle for Saipan, pp. 321-22. 

7 Unless otherwise noted, this account of the 
action of the ad Battalion, 165th, is derived from 
Love, Uattle for Saipan, pages 323-30. 

s 27th Inf Div G-3 Jnl, 25 Jun 44, Msg 38. 

Comment USNG, McDonough, Incl to Ltr, 
Brig Gen Gerard W. Kelley, to General A. C. 
Smith, n.d., OGMH. 

10 2d Bn 165th RCT Jnl, 25 Jun 44, Msg 9. 

11 165th RCT Jnl, 25 Jun 44, Msgs 51, 62. 



of his wounds, and McDonough had to be 
replaced as battalion commander by Maj, 
Gregory Brusseau. Company F was so 
stripped of its officer personnel that later 
in the day ist Lt. Joseph Trummel had to 
be transferred from G Company to take 
over command. 12 

Enemy mortars joined the machine gun, 
first to pin down and then to scatter the 
troops of Company F as they approached 
the summit of the hill. In the absence of 
any officers on the spot, ist Sgt. Edward 
Heikens took command of the company, 
collected the men, built a firm line, and 
reorganized the company front about half- 
way up the hill. There the men remained 
until late afternoon, when they were or- 
dered down the hill to dig in for the 
night. 13 

Meanwhile, Company G had reached the 
top of Hill Xray-Yoke from the south and 
was pushing out along the ridge that ran 
north from it. As it reached the point where 
McDonough and his party had been hit, 
firing broke out all along the front and 
Captain Chasmar halted his advance. The 
new battalion commander, Major Brus- 
seau, ordered Chasmar to hold fast and 
then requested tanks and self-propelled 

12 Ibid., Msg 78. 

13 As a commentary on the difficulties of terrain 
appreciation in this area of Saipan, it might be 
noted that throughout the action described above, 
battalion, regiment, and division headquarters 
were under the mistaken belief that the attack was 
being conducted not against Hill Xray-Yoke, but 
against Hill Able some 600 yards to the north (see 
27th Inf Div G-3 Jnl, Msrs 27, 28, 67). This mis- 
take derived from the erroneous reports of Cap- 
tain Chasmar of G Company, who believed himself 
to be well north of where he actually was. It was 
his report that he had reached the top of Hill 
Able that partly induced McDonough and his 
party to go exploring around Hill Xray-Yoke for 
an observation post. See Love, Battle for Saipan, 
PP- 379-8o. 

mounts. 14 However, the approach to the 
position was too precipitous and rocky to 
permit the vehicles to be brought forward, 
and when F Company withdrew Company 
G also pulled back down the hill and dug 
in with battalion for the night. 

Somewhat to the east and north of Hill 
Xray-Yoke lay another elevation, a sort of 
tongue jutting eastward from the main line 
of Purple Heart Ridge. This was labeled 
Hill Victor and lay within the zone of the 
3d Battalion, 165th Infantry. 15 Before 
jump-off hour on the 25th, Captain Betts 
of Company K had sent patrols along the 
west side of the promontory only to have 
them pinned down by fire from the top. 
Therefore he decided to move around the 
east side of the hill in an effort to take it 
from the north. Accordingly, K Company 
moved out at 0730 in column of platoons, 
2d Platoon in the lead. Immediately, the 
whole line came under heavy rifle and ma- 
chine gun fire from the top of Hill Victor 
and from caves in the sides of the cliffs. 
The company was pinned down, as was I 
Company on its right, and no further effort 
was made to reach the top of the hill. 

By midmorning it had become apparent 
that the 3d Battalion, 165th, was not mak- 
ing any progress toward reaching the top 
of Purple Heart Ridge. Colonel Kelley 
therefore decided to leave that job to the 
2d Battalion and to send the 3d off to the 
right to establish contact with the 4th Ma- 
rine Division north of Chacha. The idea 
was for the 3d Battalion to extend the right 
flank of the 106th Infantry, which was 
supposed to be moving north through the 
valley east of Purple Heart Ridge in order 

' ' ad Bn 165th RCT Jnl, 25 Jun 44, Msg 10. 

1 5 Unless otherwise noted, the account of the 
action of the 3d Battalion, 165th, is derived from 
Love, Battle for Saipan, pages 330-36. 



to seal off the Japanese in Death Valley 
from the north. 

At 1445 the 3d Battalion, 165th, moved 
off toward the north, 10 The advance was 
rapid, since there was no opposition. With- 
in a short time the battalion had estab- 
lished contact with the marines along a 
line northwest of Chacha, and by nightfall 
was digging in on the left flank of the 4th 
Marine Division. 17 To the rear, forming a 
perimeter, was the 1st Battalion, 106th, 
which had been attached to the 165th in 
the morning. 

While the 165th Infantry was making 
unsuccessful efforts to work its way to the 
top of Purple Heart Ridge, the 1st and 3d 
Battalions of the 1 06th had moved off from 
the lower end of Death Valley in the 
circling movement that was intended even- 
tually to bring them into contact with the 
2d Marine Division north of the main 
pocket of Japanese resistance. The plan, 
which had been conceived by General 
Ralph Smith and concurred in by General 
Jarman and Colonel Ayers, was for the two 
battalions to move up the valley to the east 
of Purple Heart Ridge behind the 165th 
Infantry, then cut northwest along the 
road that ran north of Hill Able and es- 
tablish a new line across the northern 
opening of Death Valley. Instead of carry- 
ing out this plan, which would have in- 
volved cutting across open country, the 
two battalion commanders, with the ap- 
proval of Colonel Ayers, chose to stick to 
the roads. The reason given was that the 
inside route was too rough to permit the 
passage of vehicles and heavy weapons. 
The decision resulted in the troops of the 
two battalions moving in a wide circle east- 

ward onto Kagman Plain into the zone of 
action of the 4th Marine Division. 18 

Colonel Ayers had ordered the 1st Bat- 
talion to make this move at 0600, but the 
troops were delayed for an hour and fifty 
minutes while the 2d Battalion, 106th In- 
fantry, extended its lines eastward in order 
to cover the position formerly held by the 
1 st Battalion and seal off the whole lower 
end of Death Valley. At 0750 the 1st Bat- 
talion, commanded by Colonel Cornctt, 
moved off, followed ten minutes later by 
the 3d Battalion under Colonel Mizony. 19 
When Colonel Cornett reached the point 
where he was supposed to turn off the road 
he was following and cut overland into the 
valley east of Purple Heart Ridge, guides 
from the 165th Infantry informed him that 
the valley route would be impassable to 
vehicles and that heavy fighting was going 
on in the area through which his battalion 
was supposed to pass. 20 Cornett then de- 
cided to continue eastward along the road 
until it crossed a road (called S Road) that 
ran northwest past Hill Able into Death 
Valley. He then intended to follow S Road 
to his assigned positions. 

Pursuing this course, the 1st Battalion 
reached the road junction, turned left, and 
had proceeded up S Road for about 400 
yards when its lead vehicles drew fire. 
There, about 1 1 30, Cornett built up a 
skirmish line 200 yards on either side of 
the road, with A Company on the right 
and B on the left. 21 Immediately upon 


1B 3 d Bn 165th RCT Jul, 25 Jun 44, Msg 43, 
17 27th Inf Div G-3 Jnl, 25 J*" 1 44, Msgs 71, 

18 Jarman, Memo for Record, 24 Jun 44, Burk- 
ner Board Rpt, Exhibit J; Gen Ross, Summary of 
Opns by 27th Inf Div, Buckner Board Rpt, Ex- 
hibit RR, p. 4. 

19 106th RCT Jnl, 25 Jun 44, p. 57. 

B " Love, Battle for Saipan, p. 337. Unless other- 
wise noted, account of the action of the 2d and 
3d Battalions, 106th Infantry, is derived from 
Love, Battle for Saipan, pages 337-45- 

21 1 06th RCT Jnl, 25 Jun 44' Msg 348. 



jumping off from this position, the battal- 
ion began to receive heavy fire from small 
arms, automatic weapons, and mortars sit- 
uated in the hills to its front. Within 
twenty minutes Company A on the right 
had .succeeded in reaching this high 
ground, although at the cost of twenty-one 
casualties, including the company com- 
mander, ist Lt. Robert C. McCoy, who 
was wounded. On reaching the high 
ground, Company A remained immobile for 
the next two hours, waiting for B Company 
on its left to come up. By 1410 Company 
B pulled abreast, and the battalion occu- 
pied a line across S Road about 400 yards 
from where it had jumped off shortly be- 
fore noon. 22 No further progress was made 
during the day. The enemy, from his po- 
sitions in the defiles of Purple Heart Ridge, 
was able to cover the whole area with con- 
tinuous machine gun and mortar fire. 
Finally, at 16 15, Colonel Ayers ordered 
Colonel Cornett to withdraw his troops to 
the road junction from which they had 
begun their movement northwest toward 
Death Valley. 23 This was accomplished by 
1840, and the ist Battalion bivouacked at 
the junction for the night. The 3d Bat- 
talion, 1 06th, meanwhile, had made no 
progress during the day beyond the same 
road junction. It had arrived there in the 
wake of the 3d Battalion, 165th, by 1155 
and reported that it was being held up by 
congestion caused by the 23d Marines. 24 
Thereafter, it made no move and finally, 
late in the afternoon, pulled back to the 
point from which it had started in the 
morning at the south end of Death 
Valley. 25 

That night General Jarman, highly dis- 
pleased with the failure of the 106th In- 
fantry to comply with its orders to skirt the 
eastern slope of Purple Heart Ridge or to 
make any significant progress along the 
wider route that it had taken, asked 
Colonel Ayers for an explanation. In the 
division commander's words, "He [Ayers] 
had no excuse and could offer no explana- 
tion of anything he did during the day. 
He stated he felt sure he could get his 
regiment in hand and forward the next 
morning (26 June) . I told him he had one 
more chance and if he did not handle his 
regiment I would relieve him." 26 

Attack up Death Valley 

The part of General Jarman's plan that 
had called for an encirclement of Death 
Valley by skirting Purple Heart Ridge to 
the east had failed. Collaterally, on 25 
June the ad Battalion, 106th Infantry, was 
to execute a holding attack across the 
mouth of the valley, contain the enemy 
there, and apply whatever pressure was 
feasible from that direction. General Jar- 
man hoped to make better use of his artil- 
lery than had been previously possible, 
and early in the morning ordered the 
commanding officer of Battery A, 106th 
Field Artillery, to conduct a reconnaissance 
along the southern end of Purple Heart 
Ridge with a view to moving one battery of 
1 55-mm. howitzers to positions from which 
they could fire directly into the cliffs that 
walled the valley on the left. 27 

By 0800 the 2d Battalion had taken its 
position across the lower end of the valley 
and by 0830 was ready to move off with 

22 Ibid., Msg 366. 

23 Ibid., p. 65, and Msg 400, 
34 Ibid., Msg 350, 

lri 27th Inf Div G-3 Jnl, 25 Jun 44, Msg 77. 

aB Jarman, Memo for Record, 24 Jun 44, Buck- 
ner Board Rpt, Exhibit J. 

27 27th Inf Div G-3 Jnl, 25 Jun 44, Msg 24. 



Lower End of Death Valley, 25 June, Elements of 106th Infantry are waiting 
for signal to attack. 

Company E on the right, G on the left. 
(Company F was still on top of the cliff, 
tied in with the 2d Marine Division.) Be- 
cause of the breadth of the front, all three 
rifle platoons of each company had to be 
committed to the line. 2s 

Though the men moved forward cau- 
tiously, the constant fire from the cliffs on 
the left precluded any real progress during 
the morning. General Jarman now, for the 
first time since the beginning of the attack 
on Death Valley, decided to bring direct 
artillery fire to bear against the cliffs on 
the left. At 1400 he ordered the 106th 
Field Artillery Battalion to move two bat- 

2X ad Bn 1 06th RGT Jnl, 25 Jun 44, Msg 27- 
Unless otherwise noted, the account of this action 
of 2d Battalion, 106th Infantry, is derived from 
Love, Battle for Saipan, pages 347-66. 

tcries of artillery into position to fire point- 
blank at the cliff line just north of Hell's 
Pocket. The artillery battalion commander 
was directed to co-ordinate his fire with 
the movement of the 2d Battalion, 106th, 
and to deliver at least a half hour's prepar- 
ation before the infantry jumped off again 
in the assault. General Jarman also at- 
tached one platoon of medium tanks from 
Company B of the 763d Tank Battalion to 
the infantry. He ordered Major O'Hara, 
commanding officer of the ad Battalion, 
1 06th, to launch his afternoon attack up 
the right side of Death Valley along Purple 
Heart Ridge using Companies E and G, 
drive all the way up the corridor, and es- 
tablish contact with Company F, 106th, 
on the left and with the 165th Infantry on 



the right. The time of attack was set for 
i6oo, a9 This drive, if successful, would by- 
pass and seal up the remaining enemy in 
Hell's Pocket and the left side of Death 
Valley and would bring the 27th Division 
abreast of the Marine divisions on either 

At 1630 the attack jumped off fol- 
lowing the artillery preparation, which was 
deemed "very satisfactory." 30 Further ar- 
tillery fire had to be called off because of 
complaints from the marines on Mount 
Tapotchau that fragments were falling 
within their lines. 31 The medium tanks 
moved into the valley ahead of the infan- 
try but were out of contact with the troops 
for the rest of the afternoon and operated 
independently, firing at will against targets 
of opportunity. 32 

Capt. David Waterson's E Company 
took the lead and was followed on its right 
rear by Company G, which was com- 
manded by Captain Tarrant, 33 The 3d 
Platoon of Company G was held in reserve 
at the battalion command post to prevent 
any possible attempt on the part of the 
enemy to break out of the valley to the 
south. Once again heavy fire from both 
sides of the valley greeted the men. The 
3d Platoon of Company E managed to 
reach the battalion's intermediate objective 
— a line of trees running across the valley 
about 800 yards from the line of departure 
— but was quickly forced to withdraw to 
the cover of another tree line about 200 

2 " 27th Inf Div G-$ Jnl, 25 Jun 44, Msg 69; ad 
Bn 1 06th RCT Jnl, a. 1 ; Jun 44, Msg 161. 

ao 37th Div Arty Unit Rpt 9, 25 Jun 44, Incl to 
27th Inf Div C1-3 Jnl, 35 Jun 44. 

S1 1 06th RCT Jnl, 25 Jun 44, Msg 394. 

' !2 Appleman, Army Tanks in the Battle for Sai- 
pan, pp. 47-48. 

ss 27th Inf Div G-3 Jnl, 25 Jun 44, Msg 84. 

yards to the rear. The 2d Platoon, Com- 
pany E, after being scattered by enemy 
fire, retreated all the way back to the line 
of departure, while the 1st Platoon never 
left it. 

Initially, Company G met with more suc- 
cess. Most of the men of this unit reached 
the tree line, which was the intermediate 
objective, and Captain Tarrant started to 
organize his defense, but a barrage from 
the side of Mount Tapotchau persuaded 
him to pull back into a gully about twenty- 
five yards to his rear. Tarrant tried to at- 
tract the attention of the tank platoon that 
was moving through the valley but neither 
smoke pots nor flares succeeded in bringing 
aid from the tanks. In the gathering dusk 
G Company, too, moved back to the first 
tree line where the 3d Platoon of Company 
E had taken cover. There, under cover of 
night, the men commenced to dig in, but 
at midnight the two company commanders 
conferred and decided to pull back to the 
line of departure. Their wounded were un- 
cared for, their ammunition, water, and 
rations were low, and they were out of ra- 
dio contact with battalion. 

The march back was full of horrors. 
Flares lighted the valley about every five 
minutes, silhouetting the retreating troops 
and occasionally revealing foxholes full of 
Japanese — who luckily did not fire. Many 
of the wounded fainted and had to be car- 
ried by their comrades; some of them died 
en route. Finally, about 0300, both com- 
panies straggled into tree line at the south 
edge of the valley that had marked their 
line of departure of the morning before, 34 
Once again the attempt to force Death 
Valley from the south had failed. 

J4 This account is derived from Love, Battle for 
Saipan, pages 353-65. 



Kagman Peninsula Secured 

In the zone of the 4th Marine Division 
the major accomplishment of the day was 
the final occupation of the whole of Kag- 
man Peninsula. This served not only to re- 
duce the corps front by about 3,000 yards, 
but also to clear the way for the construc- 
tion of an auxiliary airfield on Kagman 

The division jump-off, which was sched- 
uled at 0730, was from forty minutes to 
an hour late, but thereafter the advance 
was rapid. The 24th Marines on the right 
met little or no enemy resistance and by 
1015 had secured Kagman Hill on the 
southeastern extremity of the peninsula. 
The rest of the day was spent patrolling 
the area and investigating caves along the 
coast. On the left, the 23d Marines ran 
into occasional sniper fire and was har- 
assed by an enemy field piece located on 
Purple Heart Ridge but nevertheless man- 
aged to reach the O-fi line on the cast 
coast by 1533. Thus Kagman Peninsula 
was completely blanketed and sealed off, 
and the 4th Marine Division for the first 
time in four days was permitted to relax. 35 

Seizure of Mount Tapotchau 

Honors for the capture of the summit of 
Tapotchau, the highest point of the island, 
were shared by the 2d Battalion, 8th Ma- 
rines, and the 1st Battalion, 29th Ma- 
rines. 36 The 2d Battalion, 8th Marines, 

35 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RGT Rpt, pp. 41-42; Annex I, 24th RCT 
Rpt, p. 31. 

30 The following account of the actions of the 
ad Marine Division is derived from: 2d Marine 
Div SAR, Phase I, Forager, Sec. VI, p. 14; and 
Hoffman, Saipan, pp. 151-55. 

pushed off at 0730 and advanced rapidly 
against little resistance along the cliff line 
overlooking Death Valley until it found it- 
self, three hours later, at the base of a 
sheer fjoToot cliff, just beyond which was 
the peak of the mountain itself. Patrols 
scaled the cliff, worked their way almost to 
the crest of the mountain, and returned 
shortly after noon with the information 
that the small plateau that constituted the 
summit of Tapotchau was unoccupied. On 
receiving this information, the 1st Battalion 
of the 29th Marines, on the left, gradually 
worked its way to the right along the route 
that had been followed by the patrols and 
by late afternoon had established itself on 
the mountain top. During the operation 
enemy fire was not severe, but that night 
a force of Japanese counterattacked and 
had to be repulsed. Eighteen enemy dead 
were counted the next morning. During 
this period Company F, 106th Infantry, 
moved forward with the 2d Battalion, 8th 
Marines, guarding its right rear flank. 

On the western slopes of Tapotchau the 
other two battalions of the 8th Marines 
made little progress against the multitude 
of machine gun nests in the network of 
crevices and ravines that crisscrossed that 
side of the mountain. The heaviest fighting 
occurred on the right flank of the 6th Ma- 
rines, where for the third successive day 
the attack was held up by the Japanese 
strongpoint north of Mount Tipo Pale. 

On the left of the 6th Marines' zone the 
lines had been extended as far as was feasi- 
ble, and no further advance was possible 
until the difficulty on the right had been 
cleared up. On the western coast and on 
the high ground overlooking Garapan the 
2d Marines, too, remained immobile, wait- 
ing for the elements on its right to come 



In terms of yardage gained the 2d Ma- 
rine Division had made little progress on 
25 June, but the capture of Tapotchau's 
peak yielded it control of Saipan's key ter- 
rain feature. In the words of the official 
Marine Corps historian, "From this point 
forward, the Marines would look down on 
the Japanese, who heretofore had enjoyed 
the advantages of superior ground observa- 
tion. In addition the Marines could now 
fight downhill for a while. The change 
would be appreciated." 37 

The Plight of the Japanese 

By the end of 25 June it was obvious 
to the Japanese high command on Saipan 
that the situation was desperate almost 
(but not quite) to the point of hopeless- 
ness. A telegram from gist Army head- 
quarters to the 29th Division on Guam 
stated that the ten-day battle had reduced 
the strength of the line forces to the follow- 
ing approximate levels: 38 

Unit Strength 

1 1 8th Infantry Regiment 30Q 

135th Infantry Regiment (exclusive 
of the 1 st Battalion, which was 
on Tinian) 350 

136th Infantry Regiment 300 

47th Independent Mixed Brigade 100 

yth Independent Engineers 70 

3d Independent Mountain Artillery 

Regiment (no weapons 

in firing 
Qth Tank Regiment 3 tanks 

In addition to the losses, the message 
added, "reserve units (companies and pla- 
toons), hospital units, equipment, mainte- 
nance and supply units, etc.. are either 
completely wiped out or reduced to the 
point where no fighting strength can be 
expected of them," 

37 Hoffman, Saipan, p. 153. 

■ 1 ' 1 31st Army Msg File, Msg 1 102. 

With the failure of the defense effort 
along the line south of Tapotchau the Jap- 
anese, in spite of occasional lapses of wish- 
ful thinking, realized full well that the 
island could not be held. In a lengthy tele- 
gram, probably of 25 June, 31st Army 
headquarters said as much and analyzed 
some of the reasons for its failure: 

The fight on Saipan as things stand now 
is progressing one-sidedly since, along with 
the tremendous power of his barrages, the 
enemy holds control of sea and air. In day- 
time even the deployment of units is very 
difficult, and at night the enemy can make 
out our movements with ease by using illumi- 
nation shells. Moreover, our communications 
are becoming disrupted, and liaison is be- 
coming increasingly difficult. Due to our 
serious lack of weapons and equipment, ac- 
tivity and control is hindered considerably. 
Moreover, we are menaced by brazenly low- 
flying planes, and the enemy blasts at us 
from all sides with fierce naval and artillery 
cross-fire. As a result even if we remove units 
from the front lines and send them to the 
rear their fighting strength is cut down every 
day. Also the enemy attacks with fierce con- 
centration of bombs and artillery. Step by 
step he comes toward us and concentrates 
his fire on us as we withdraw, so that wher- 
ever we go we're quickly surrounded by 
fire. 39 

Continuing, the message noted two dif- 
ficulties peculiar to the Saipan campaign. 
The first was the confusion caused by the 
presence on the island of so many strag- 
gler units, the waifs cast up by American 
submarine attacks. The second was the 
ever-growing shortage of water. There had 
been little enough at the beginning of the 
campaign, and the American bombard- 
ment had closed many of the sources of 

The prospect was dim: "The attack of 
the enemy proceeds ceaselessly day and 

3V Ibid,, Msg 1 101. 

MAP 9 



night and as they advance with the aid of 
terrific bombardments it becomes apparent 
that the northern part of the island for 
the above mentioned reasons of ( i ) water, 
(2) food, (3) supply, and (4) terrain, can- 
not be held with our skeleton strength of 

Yet in spite of the admitted futility of 
resistance, resistance continued. Surrender, 
the only practical thing to do in such a 
situation by Western standards, was out of 
the question for the Japanese. The order of 
the day read, ", . . the positions are to be 
defended to the bitter end, and unless he 
has other orders every soldier must stand 
his ground." 40 

Grasping at straws, General Saito on 24 
June had ordered an infantry company 
from Tinian to conduct a landing opera- 
tion on the coast of Saipan, east of 
Chacha. 41 On the night of 25 June eleven 
barges departed Sunharon ( Tinian ) harbor 
for Saipan. The destroyer Bancroft inter- 
cepted and dispersed them. One of the 
barges was reported sunk; the rest scurried 
back to Tinian Town. Still later, in the 
early morning hours of 26 June, several 
troop-laden barges came out of Tanapag 
Harbor, destination unknown. Two LCI 
(G)'s on patrol opened fire, sank one of 
the barges, and damaged another. 42 Thus 
ended Saito's immediate prospects of aid 
from counterlandings. 

26 June 
Action of the 27 th Division 

With the failure of the 106th Infantry 
to accomplish its mission on the 25th, Gen- 

10 NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, p. 104. 

41 31st Army Msg File, Msg 1093. 

4= TF 51 Opn Rpt Marianas, Incl A, p. 9. 

cral Jarman proposed a new scheme of 
maneuver for that regiment on the 26th. 
His plan called for the 3d Battalion to push 
along the inside (western) slope of Purple 
Heart Ridge, build up a line there, and, if 
possible, push on to the regimental objec- 
tive line at the north end of Death Valley. 
The 2d Battalion, after reorganizing, was 
to follow the 3d, then swing left across the 
valley and move on to the regimental ob- 
jective. The 1 st Battalion would be in regi- 
mental reserve. Later (at 0920) Colonel 
Ayers ordered the 1st Battalion to move 
out on the left of the valley and clean out 
Hell's Pocket, which was still infested with 
Japanese. 43 At the same time the 2d Bat- 
talion, 165th, was to continue the attack 
against P urple Hea rt Ridge from the 
southeast. 44 1 {Map g) \ 

The 3d Battalion, 106th, jumped off at 
0600 in column of companies with Com- 
pany L in the lead, followed by I, M, and 
K. 45 By 1020 the leading elements of 
Company L had reached the top of Hill 
Oboe without encountering significant re- 
sistance and had started down into the 
saddle between Oboe and Hill King to the 
north. 46 Company I followed immediately 
behind. As Company L advanced toward 
Hill King a machine gun opened up, and 
the heavy weapons company was called 
up to train its mortars and machine guns 
on the suspected source of fire. At 1245, 
under cover of this protection, Company 
L moved forward, but within ten minutes 
the lead platoon had six men killed and 
seventeen wounded. 47 The advance halted, 

13 106th RCT Jnl, s6 Jun 44, p. 70; 106th Inf 
Narrative Rpt Forager, p. 9. 

44 Unless otherwise noted, the account of the 
actions of the 27th Division on 26 June is derived 
from Love, Battle for Saipan, pages 375^96. 

45 27th Inf Div G-3 Jnl, 26 Jun 44, Msg 6. 
40 3d Bn 1 06th RCT Jnl, 26 Jun 44, Msg 16. 
47 27th Inf Div G-3 Jnl, 26 Jun 44, Msg 44. 



Bazooka Team Preparing To Fire a 2.36-inch rocked launcher during mopping- 
up operations along the ridge line. 

then the company was withdrawn. By the 
time Captain Hallden had worked his way 
back to Hill Oboe and reorganized his 
company, it was in a highly demoralized 
condition. The 1st Platoon was down to 
twelve men, as was the 2d Platoon, which 
had been hard hit during the explosion of 
the enemy ammunition dump on 23 June. 
Company I was sent in to relieve Company 
L. It enjoyed no more success than its pre- 
decessor and retired to Hill Oboe to dig in 
with the battalion for the night. 

The 2d Battalion, 106th, had stayed be- 
hind the 3d during the day and made no 
effort to work its way out into the valley. 
It dug in on Hill Xray-Yoke for the night. 

While the 106th Infantry was moving 
along the inside of Purple Heart Ridge, the 

2d Battalion, 165th, had begun to mop up 
on the outside (east) of the ridge line. 
Major Brusseau had ordered Company G 
to fan out and clean up Hills Xray-Yoke, 
Oboe, and Victor. Upon completion of this 
mission Company G was to move on north 
to Hill Able. Meanwhile, Company E was 
to push patrols to the base of Hill Victor 
and then proceed up S Road to the point 
where it cut into Death Valley. 48 

The 1 st Platoon, Company G, took over 
Hill Xray-Yoke without opposition. By 
0840 Captain Chasmar's 3d Platoon was 
atop Hill Victor without much trouble, but 
thereafter the Japanese began to show 
some fight. The enemy had taken refuge 

48 Love, Battle for Saipan, p. 379. 



in ledges and caves just below the crest of 
the hill and from there commenced lob- 
bing hand grenades up into the line of the 
Americans on top. Some thought was 
given to getting at these positions by tying 
charges onto ropes and letting them swing 
down into the protected strongpoints, but 
rather than attempt this device, Chasmar 
withdrew the lead platoon to the ground 
below and called up self-propelled mounts 
(M7's) to fire into the ledges where the 
enemy was entrenched. The M7's failed to 
accomplish the mission and had to retreat 
when the Japanese began dropping mortar 
shells on them. Three tanks were then 
brought forward, but to no avail. As a last 
resort Captain Chasmar sent up some M8 
self-propelled mounts. These vehicles were 
equipped with 75-mm. rather than 105- 
mm. howitzers and had smaller openings at 
the top, thus offering better protection to 
the gunners from fire from above. For half 
an hour the M8's plugged away at the sides 
of the cliffs, forcing many of the enemy 
into the open where they could be picked 
off by riflemen. 

At 1500 the 3d Platoon ventured to the 
top of the hill again, followed an hour later 
by the 2d Platoon. Machine gun fire from 
the west along the main line of Purple 
Heart Ridge held down their advance how- 
ever, and since darkness was approaching 
they were recalled down the hill to dig in 
for the night. 49 

Meanwhile, E Company had worked its 
way under scattered fire up S Road, along 
the route taken the day before by the 1st 
Battalion, 106th Infantry. It had reached 
the point of the previous day's farthest ad- 
vance and had begun to dig in when the 
Japanese on the hills to the left opened 

4 * Ibid,, pp. 388-91. 

with machine guns and mortars. Inas- 
much as the company was in an isolated 
position almost a thousand yards from the 
rest of the battalion, the new commander, 
1st Lt. John J. Raleigh, took his men back 
to join the rest of the battalion for the 

Purple Heart Ridge was beginning to 
crack under the combined assault of the 
2d Battalion, 165th, and the 3d Battalion, 
1 06th Infantry. Meanwhile, on the other 
side of Death Valley, Colonel Cornett's 1st 
Battalion, 1 06th, was vainly trying to clear 
out Hell's Pocket. The battalion moved off 
at 1245 following a thirty-minute prepara- 
tion by the Cannon Company but was 
quickly pinned down by fire from the 
pocket. 51 Company C alone lost three men 
killed, and twenty-two wounded including 
Capt. Robert T. Bates, the company com- 
mander, who was replaced by 1st Lt. An- 
drew B. Campbell. Within the next hour 
the 104th Field Artillery put 360 rounds 
of 105-mm. howitzer fire into the pocket. 
It was to no avail. The battalion still en- 
countered heavy fire from the walls of the 
pocket and from the high rock in the 
center, so Cornett retired his men to the 
line of departure where they dug in for the 
night. 32 

Sometime after ten o'clock on the morn- 
ing of 26 June General Kcrnan, division 
artillery commander, in company with 
Colonel Sheldon, the operations officer, 
made a trip up to the southern edge of 
Death Valley. There they found that the 
2d Battalion had made no advance out of 
the assembly area, where it was mingled 
with the rear elements of the 3d Battalion. 

50 Ibid,, p. 39a. 

sl 27th Inf Div G-3 Jnl, a6 Jun 44, MsgS 39, 

53 Love, Battle for Saipan, pp. 393-94. 



They reported that "the battalions were 
standing still and there was no reason why 
they should not move forward." The to6th 
Infantry, they concluded, was in a demor- 
alized state. On the basis of this report, 
which confirmed his previous dissatisfac- 
tion with this regiment's conduct, General 
Jarman relieved Colonel Ayers of his com- 
mand and assigned Colonel Stebbins, the 
division chief of staff, as commander. 5,1 

Action of the Marines 

Having completed its occupation of 
Kagman Peninsula, the 4th Marine Divi- 
sion was ordered on the 26th to mop up 
the area, outpost the coast line of Magi- 
cienne Bay, and then assemble in the 
vicinity of the beaches along the northern 
coast of the bay in corps reserve. The ma- 
rines encountered no enemy opposition 
except from small groups of snipers in the 
vicinity of Chacha and in the caves along 
the coast, although Japanese artillery oc- 
casionally opened up from the unsecured 
portions of Purple Heart Ridge. Before the 
division could properly assemble in corps 
reserve plans were changed, and it was or- 
dered to take over the right of the line 
again next morning. 

In the zone of the 2d Marine Division 
the most important event of the day was 
the bypassing of the pocket north of Mount 
Tipo Pale by the 2d Battalion, 6th Ma- 
rines. One company passed to the right, 
another to the left, while the third was left 
behind to reduce the pocket. 

The 8th Marines registered only small 
gains in the Mount Tapotchau area. In the 

03 Jarman, Memo for Record, 24 Jun 44, Buck- 
ner Board Rpt, Exhibit J; 27th Inf Div Special 
Order 118, 26 Jun 44, Inc.l to 27th Inf Div G-3 

54 4th Marine Div Opns Rpt Saipan, p. 27. 

regimental center the 3d Battalion, 8th 
Marines, and the 1st Battalion, 29th Ma- 
rines, moved forward slowly through night- 
marish terrain, receiving heavy mortar and 
machine gun fire as they went. The 2d 
Battalion, 8th Marines, and Company F, 
to6th Infantry, remained stationary on the 
right flank, holding back to deny enemy 
approaches from that direction. No great 
yardage was gained in this area, but the 
positions on the heights of Tapotchau were 
consolidated and the regimental line was 
straightened out. On the division left the 
2d Marines again remained immobile ex- 
cept for patrols, who reported that there 
was no Japanese activity to the immediate 
front. 55 

27 June 
Death Valley Broached 

The first permanent inroads into Death 
Valley were made on 27 June. Up to that 
date the only significant progress in the 
zone of the 27th Division had been in the 
hills that made up the lower part of Purple 
Heart Ridge. Death Valley itself had defied 
capture; the Japanese from their com- 
manding positions in the cliffs on the left 
and the northern part of Purple Heart 
Ridge on the right were able to interdict 
any movement along the floor of the valley 

General Jarman's plan for 27 June called 
for a reorientation of the direction of the 
attack. For this purpose he had four bat- 
talions under his control — all of the 106th 
Infantry plus the 2d Battalion, 165th, 
which had been detached from its parent 
regiment the night before when the 165th 

50 ad Marine Div SAR, Phase I, Forager, Sec. 
VI, p. 15- 



Regiment was attached to the 4th Marine 
Division. 50 The 2d Battalion, 165th, was 
ordered to continue mopping-up opera- 
tions against Purple Heart Ridge, working 
from the eastern slopes. Meanwhile, the 3d 
Battalion, 106th, was to move north along 
the ridge until it reached Hill King, then 
pivot left and attack west across the valley 
toward Mount Tapotchau. Once a corri- 
dor had been driven across the valley, the 
battalion was to pivot again, this time to 
the right, and move off toward the north- 
ern end of the valley. Before this drive was 
completed the 2d Battalion, 106th, was to 
take positions on the right flank of the 3d. 
In the meantime the 1st Battalion, 106th, 
in an independent movement, would 
renew its attempt to clean out Hell's 
Pocket. 57 

The 3d Battalion, 106th, jumped off for 
Hill King at 0620, with Company I on the 
right, K on the left. 58 Almost immediately 
machine gun fire opened up, killing one 
man and wounding seven, and the bat- 
talion was ordered back to Hill Oboe, 
which had been the line of departure. 
Division then ordered twenty-five-minute 
artillery preparation, to commence at 1 020, 
but the position of the American troops on 
Purple Heart Ridge was so hard to ascer- 
tain that the artillerymen held fire for more 
than half an hour. 50 

Following the artillery preparation, 
which was completed by 11 20, the attack 

r,r ' 27th Inf Div Fid Msg i(a), 26 Jun 44, Incl 
to 27th Inf Div G-3 Jnl. 

57 Love, Battle for Saipan, pp. 400-401, 403, 


C8 27th Inf Div G-3 Jnl, 27 Jun 44, Msg 24. 
Unless otherwise noted, the account of the action 
of the 3d Battalion, rofith Infantry, is derived 
from Love, Battle for Saipan, pages 401-03, 

Si ' 27th Inf Div G-3 Jnl, 27 Jun 44, Msgs 33, 

moved off again with Company L on the 
left in place of Company K, whose strength 
was now down to about that of one pla- 
toon. This time there was no opposition. 
The battalion moved up Hill King through 
a Utter of enemy dead, and not a shot 
was fired. As Company I moved over the 
crest of the hill and down its northern 
slope, it surprised a large party of Japanese 
hiding among the rocks and grass. After a 
brief exchange of rifle fire and hand gren- 
ades, the Americans withdrew to the re- 
verse slope and mortar fire was requested. 
This lasted only a few minutes, after which 
the attackers were able to push down the 
north slope of the hill without trouble. 

With Hill King secured, the 3d Battal- 
ion, 1 06th Infantry, could now push off 
to the west across Death Valley according 
to plan. The attack jumped off at 1150, 
Company L on the left and I on the 
right/' The terrain to the battalion front 
was a steep slope down into the valley 
proper. The descent would be made 
through thick, high tufts of grass for most 
of the way, and then through cane fields 
into a low ridge line that cut across the 
valley at that juncture. Directly to the bat- 
talion front, about a thousand yards away, 
were the cliffs of Mount Tapotchau, 

Immediately upon reaching the floor of 
the valley, the men of the 3d Battalion be- 
gan to receive machine gun and mortar 
fire from Hill Able behind them and from 
the cliff sides of Tapotchau to their front. 
Among others wounded was 1st I-t. Robert 
M. Smith, who had taken command of 
Company I only that morning. This left 
the company without officers except for 
one platoon commander, whose unit had 
become separated from the rest of the 

«° 1 06th RCT Jnl, 27 Jun 44, Msg 545- 



company. In the hiatus thus created, Cap- 
tain Hallden of Company L incorporated 
the scattered remnants of I Company with 
his own unit and took command of both. 

Meanwhile, Company K had been or- 
dered into the line on L Company's left, 
with the mission of securing the tree line 
that ran across the valley at this point and 
of establishing physical contact with Com- 
pany F, 1 06th Infantry, which was still on 
the cliff top just below the summit of 
Mount Tapotchau. No sooner was Com- 
pany K abreast of L Company than it, too, 
came under heavy fire that by now was 
general throughout the valley floor. 

Nevertheless, the 3d Battalion succeeded 
in cutting across the valley and was send- 
ing out patrols to establish contact with 
Company F on the clifT in front by 1545. 61 
By this time all three companies were badly 
in need of ammunition. They had no sup- 
plies of water or rations, and parts of each 
company had been cut off from the main 
body of the battalion. Colonel Mizony 
placed Captain Hallden in charge of the 
remnants of all three companies and dis- 
patched a platoon of light tanks, not only 
to lay fire on both sides of the valley but 
to supply the infantry with rations, water, 
and ammunition/' 2 The tanks accom- 
plished their mission before dark, and the 
3d Battalion dug in along the low ridge 
line that traversed Death Valley west of 
Hill Able. There it was joined by the 2d 
Battalion, which had come up behind, and 
the two battalions prepared to attack to 
the north the next morning. 

Meanwhile at 11 20, following a delayed 
artillery preparation, Company G, 165th 

Infantry, pushed up the eastern slope of 
Hill King, which 3d Battalion, 106th In- 
fantry, was attacking from the other 
sidc. uS Within three hours G Company 
had reached the summit of the hill and was 
driving down its forward slope toward Hill 
Able to the north. Hill Able was more an 
outsized rock than a hill. The south face 
toward which Captain Chasmar's company 
was moving was a sheer cliff about fifty 
feet high, crosscut by a series of ledges. The 
right (east) face of the hill was steeply- 
terraced and the west side was another 
sheer cliff. The top was a rounded knob 
covered with dense undergrowth and was 
heavily defended, judging from the intense 
fire that began to pour down on the heads 
of the advancing troops. 

About 1600, before it reached the foot 
of the hill, Company G was attacked by a 
party of Japanese moving down through 
the corridor between Hills King and Able. 
The enemy movement was undetected un- 
til the Japanese were within a few yards 
of Chasmar's positions. A severe hand-to- 
hand fight ensued, resulting in seven 
American casualties and thirty-five enemy 
killed. By this time it was too late to 
warrant a continuation of the attack 
against Hill Able, and G Company moved 
back to Hill King to dig in with the bat- 
talion for the night. Before the men could 
prepare their foxholes, however, a heavy 
barrage of mortar and machine gun fire 
fell on the area, killing five and wounding 
nineteen. Among the latter was Major 
Brusseau, the 2d Battalion commander, 

01 27th Inf Div G-3 Jul, 27 Jun 44, Msg 71. 
"- Applcman, Army Tanks in the Battle for Sai- 
pan, pp. 51-53- 

' i;l 27th Inf Div G-3 Jnl, 27 Jun 44, Msg 39; 
Love, Battle for Saipan, pp. 419-27. Unlr-ss other- 
wise noted, this account of the actions of the ad 
Battalion, 165th Infantry, is derived from the lat- 
ter source. 



who later died of his wounds. He was 
replaced by Capt. James A. Dooley, 6 ' 1 who 
was in turn later relieved by Major Claire. 
Claire's command of the 3d Battalion, 
165th, was taken over by his executive of- 
ficer, Maj. Martin Focry. At the conclusion 
of the fire G Company, 165th Infantry, 
withdrew past Hill Oboe and back to Hill 
Xray-Yoke for the night, where it was 
joined by Company E. The latter had 
spent the day unsuccessfully trying to 
move up S Road to the point where it 
entered Death Valley. Scattered rifle fire, 
coupled with heavy mortar and machine 
gun fire from a hill that commanded the 
road, had thwarted the effort to break into 
Death Valley by this route. 

During the afternoon General Jarman 
had still been skeptical of the staying power 
of the 106th Infantry and had instructed 
the executive officer of that regiment to get 
word to all units that "they" must hold and 
under no case fall back." en Now, with the 
drive across Death Valley successfully com- 
pleted, Jarman was relieved and gratified. 
To the commanders of the 2d and 3d Bat- 
talions he dispatched the message: "Con- 
gratulations on a day's work well done. 
I have the utmost confidence in our 
continued success in a vigorous push 
against the remaining enemy. Keep up the 
good work." fifi 

In the meantime, the 1st Battalion, 
106th Infantry, was taking the first effec- 
tive step of the campaign toward mopping 
up Hell's Pocket at the southwest end of 
the valley. To accomplish this mission, 
Colonel Cornett planned to use Company 
C to contain the Japanese at the mouth of 

27th Inf Div G-3 Jnl, 27 Jun 44, Msg 107. 

Ibid., Msg 87. 

1 06th RCT Opns Rpt Forager, p. 1 1. 

the pocket while the other two rifle com- 
panies climbed the cliff to the left in order 
to approach the enemy's positions from 
above and the rear/' 7 

Company A took the lead in the envelop- 
ing movement, slowly groped its way to the 
top of the cliff, and commenced circling the 
rim around Hell's Pocket, Soon the lead 
platoon stumbled upon a deep crater, al- 
most fifty yards wide, that turned out to be 
a nest of live and very active Japanese. 
Grenades and rifle fire failed to silence the 
position. Mortars were then dragged up 
the cliff to accomplish the job, after which 
the infantrymen moved on through and 
past the crater. Twenty dead Japanese 
were found in the area, as well as two ma- 
chine guns and three fully operative Amer- 
ican Browning automatic rifles. 6 " Com- 
pany B, meanwhile, was held up behind A 
Company until the latter had cleaned out 
the crater. During this wait the battalion 
executive officer, Maj. John Nichols, who 
was in charge of the cliff-top operation, 
came forward and relieved 1st Lt. Frank J. 
Pryor of command of the company replac- 
ing him with 1st Lt. Charles Warge. The 
new company commander immediately be- 
gan to deploy his unit to the left, and by 
the time the crater was cleaned out he was 
abreast of Company A, Both companies 
then formed a skirmish line and moved 
forward another hundred yards north 
along the edge of the pocket without flush- 
ing any more Japanese. There, Major 
Nichols ordered the advance halted and 
both companies pulled 500 yards to the 
left of the cliff line where they dug in for 
the night. 

V7 Ibid. Unless otherwise noted, the account, of 
the action of 1st Battalion, 106th Infantry, is de- 
rived from Love, Battle for Saipan, pages 428-34. 

fis 106th RCT Opus Rpt Forager, p. n. 



Action on the Flanks 

On the right flank of the three-division 
front, the 4th Marine Division was ordered 
to continue northward and seize all of the 
0-6 line witlrn its zone. 00 For this opera- 
tion the division had under its control a 
total of nine battalions — the 23d and 24th 
Marines and the 165th Infantry (less 2d 
Battalion) with the 1st Battalion, 105th 
Infantry, attached. 

On the right, the 23d Marines made 
rapid progress against only occasional 
small arms fire from the villages of Hashi- 
goru and Donnay and by 1640 had 
reached its objective. The advance would 
have been even more rapid but for dense 
underbrush and the ragged cliff line along 
the coast that called for cautious move- 
ment and thorough investigation. 70 

On the left of the 23d Marines, the 
165 th Infantry jumped off at 0730 on 
schedule with the 3d Battalion on the 
right, in direct contact with the left flank 
of the 23d Marines, and the 1st Battalion, 
105th Infantry, on the left, followed about 
600 yards to the rear by the 1st Battalion, 
165th. 71 The advance was rapid, especially 
on the right flank where the 3d Battalion, 
165th, met little resistance. Toward late 
afternoon the 1st Battalion, 105th, encoun- 
tered some heavy enemy fire from positions 
along the road that ran west of the village 
of Hashigoru. Colonel Kelley ordered the 
battalion commander to bypass the posi- 
tions, which he did. 

69 NTLF Opn Order 14-44, 26 Jun 44. 

10 4th Marine Div Opns Rpt Saipan, Annex H, 
03d Marines Rpt, p. 44, 

71 165th RGT Jnl, 27 Jun 44, Msg 29 ; 165th 
RCT Rpt of Action Saipan, p. 6. Unless otherwise 
noted, this account of the action of the 165th 
Infantry is derived from the latter source. 

Meanwhile the ist Battalion, 165th, 
which was moving up on the left rear flank 
of the regimental line, had lost all contact 
with the 2d Battalion, 165th, which was 
still held up among the hills of Purple 
Heart Ridge. On Colonel Kclley's recom- 
mendation the 3d Battalion, 24th Marines, 
was brought up and committed to fill the 
gap. Action in the regimental zone was 
broken off about 1700, and all three bat- 
talions of the 165th Infantry dug in for 
the night in positions that bent back on 
the left to retain contact with the 106th 

The 2d Marine Division's advance on the 
27th was much slower than that of its sis- 
ter division because of continued terrain 
difficulties around Tapotchau and heavy 
Japanese resistance in the center. Dur- 
ing the morning the marines completed the 
occupation of the entire main crest of 
Mount Tapotchau and started down its 
northern slope. 72 In the division center 
the lines had been tightened sufficiently by 
noon to permit one badly battered battal- 
ion of the 6th Marines to be pinched out 
and retired to reserve status. Later that 
afternoon the battalion on the 6th Marines' 
right ran into heavy resistance from the 
ridge line north of Tipo Pale and was 
stopped in its tracks, causing the remainder 
of the regiment to hold back too, although 
the opposition on the regimental left had 
been comparatively light. On the extreme 
left flank the 2d Marines was again com- 
pelled to stand still on the outskirts of Gara- 
pan until the rest of the division pulled 
abreast. It spent the day consolidating its 
positions and sending out patrols. 

7S 2d Marine Div SAR, Phase I, Forager, Sec. 
VI, p. 16; Hoffman, Saipan, pp. 1 7 1-73. 



Japanese Reactions 

June 27 th marked another turning point 
for the Japanese in their stubborn, futile 
battle to save Saipan from the invaders. 
On that day General Saito established a 
final line of resistance where a last stand 
would be made. This was the third such 
battle line to be laid down. The first had 
been the shore line; the second, the mid- 
island defense line ; and now a third was to 
be held across the island from Tanapag on 
the west coast, through Hill 221 (meters) 
and Tarahoho to the east coast, cutting 
across the ba se of the island's northern tip. 
I (See Map 7.71 

The withdrawal was to be gradual. 
What remained of the mid-island defense 
line would be held until the new line could 
be established. This policy of conducting 
a delaying action until a new line could be 
built up was explained by gist Army head- 
quarters to Tokyo: "The Defense Force, 
along with the firmest possible defense of 
its present defense line and its activities 
toward annihilation of the enemy, is at 
present setting up with a line between 
Tanapag-Hill 221-Tarahoho as the final 
line of resistance." 73 

General Saito still clung to the small 
hope that a renewed Japanese air attack 
might alter the situation and save Saipan 
and the Marianas for the Empire. In a 
telegram to the assistant chief of staff in 
Tokyo, to the Minister of War, and to the 
2gth Division on Guam, he painted a 
gloomy picture of the future of Japanese 
forces on Saipan, but closed with a hopeful 
reference to the Marpi Point airfield, 
which was still in his hands: 

The pressing need of the moment is that 
the mistake be not made of allowing this 

7;i 31st Army Msg File, Msg 1 120. 

important experience in the defense of Sai- 
pan to be put to no practical end, and, the 
soldiers here be robbed of the fruits of victory 
after having fought so bravely. . . . Especi- 
ally, the Eanadero [Marpi Point] airport 
has not been completed, but in case the ne- 
cessity arises, it can be used, and the Saipan 
defense forces trust that they can hold out 
until the first 10 days of the month [July], 
awaiting its completion. 74 

In another message to Guam, the Japa- 
nese general made a more specific plea for 
air reinforcements to be sent to Tinian: 

The attacking force of the enemy has the 
appearance of becoming less intense from 
now on. Even though the Banadero airport 
has not yet been completed, we are endeavor- 
ing to finish it, so that it may be an air base 
in the Marianas which we can use. However, 
for the present it is an impossibility. 

Because the enemy planes which have ap- 
peared in the air are only carrier borne 
bombers and recco planes, the situation is 
such that our large fighter formations could 
seize good opportunity for daylight sinking 
of enemy destroyers, etc. 

However as the fate of the Empire will be 
decided in this one section, we trust that you 
will decide to send fighters to Tinian. 7C 

28 June 
Action of the 27th Division 

The key to the battle in the zone of the 
27th Division on 28 June was the fight for 
Hill Able, the northernmost promontory of 
Purple Heart Ridge, and the failure of the 
2d Battalion, 165th Infantry, to capture 
this hill brought progress through Death 
Valley to a standstill once more. After the 
heavy pounding that had been taken by G 
Company on the preceding night, the 
whole battalion had pulled back from Hill 
King, past Hill Oboe, and had dug in atop 

14 Ibid., Msg 1121. 
7fi Ibid., Msg 1 122. 



Hill Xray-Yoke in almost the same place 
where it had spent the previous two nights. 
On learning of this, General Jarman ex- 
pressly ordered, "No [future] withdrawal 
will be made for the night for the purpose 
of consolidation." 7S 

Major Claire had been transferred the 
night before from the 3d Battalion, 165th, 
to take over command of the 2d Battalion 
after Major Brusseau had been mortally 
wounded. Claire's plan for the 28th called 
for G Company again to make the assault 
along the ridge. F Company was to circle 
the low ground on the east side and come 
up between Hills King and Able. Company 
E was to be main tained as battalion re- 
serve. 77 f^7a/^^ 

Company G jumped off promptly at 
0630 and pushed rapidly over Hill Oboe. 
On reaching Hill King the men discovered 
that the enemy had either rcoccupied the 
hill during the night or had remained well 
concealed from the American troops who 
had "captured" it the previous day. At any 
rate, the Japanese suddenly came to life 
with machine guns and rifles and the ad- 
vance of Company G was stopped. Around 
noon, self-propelled mounts were brought 
forward and after an hour's fire from these 
vehicles enemy fire ceased. Company G 
moved forward again and encountered no 
trouble until the men went over the crest 
of the hill. There, the whole line was 
greeted by a shower of grenades and ma- 
chine gun fire from the east slope, which 
had not been touched by the self-propelled 

For half an hour there was a furious fire 
fight, but the Americans did not advance. 
At 1330 Captain Chasmar called battalion 
headquarters to report his casualties, 
which numbered about twenty. 78 He was 
ordered to pull back to Hill Oboe until 
mortar fire could be brought to bear upon 
the enemy line. At Oboe he stopped to re- 
organize, but enemy mortar shells falling 
into his lines caused complete confusion. 
Major Claire ordered one platoon of Com- 
pany E to move up and take over Hill 
Oboe. Company G was withdrawn to Hill 
Xray-Yoke, where it dug in in the bivouac 
area of the night before. There it was 
joined by Company F, which had met with 
no more success in trying to assault Hill 
King from the east. 

The setback suffered by the 2d Battal- 
ion, 165th Infantry, on the 28th was to 
govern, retard, and finally frustrate the ef- 
fort of the 2d and 3d Battalions of the 
106th to break out of Death Valley. So long 
as the Japanese held Hill Able they could 
interdict the entire northern half of the 
valley and pour devastating fire on any 
troop movements through the area. 

The attack order of the two battalions 
of the 106th Infantry called for them to 
jump off at 0630 on 28 June, following a 
fifteen-minute artillery preparation, and to 
capture the 0-6 line, which lay about 
3,000 yards to the north, 79 Between them 
and their objective lay a series of low ridge 
lines, the first one of which crossed the val- 
ley about 400 yards to their front. 

The advance moved off on schedule with 
the 2d Battalion on the right and the 3d on 
the left, and by 0930 the men had reached 

7f ' 27th Inf Div FO 51, 37 Jun 44, Incl to 5>7th 
Inf Div G-3 Jul, 27 Jim 44. 

77 Unless otherwise noted, the account of the 
action of the ad Battalion, 165th Infantry, is de- 
rived from Love, Battle for Saipan, pages 456-63. 

'« 3d Bn 165th RCT Jnl, 28 Jun 44, Msg 30. 

7!) 27th Inf Div FO 51, 27 Jun 44, and 106th 
Inf FO ro, 27 Jun 44., Incls to 27th Inf Div G-3 
Jnl, 28 Jun 44. 


28-30 June 1944 

immtffl!}!! Front line, Evening 28 JuNt 
00000 Positions rciched 29 June 
IH'tllTH Positions reached 30 June 

MAP 10 



the first ridge line. 80 There they were or- 
dered to dig in to await supplies that were 
to be brought forward in self-propelled 
mounts. Almost as soon as the vehicles ap- 
peared and began to unload, the enemy, 
who had been more or less quiescent for 
over an hour, opened up from Hill Able 
with intense fire. The self-propelled mounts 
promptly dropped their supplies and scur- 
ried for cover, 81 and the infantry com- 
manders had to send out carrying parties 
to pick up the supplies. The men who went 
back to recover the hastily jettisoned sup- 
plies were caught in heavy enemy mortar 
fire that was being directed at the Cannon 
Company vehicles. Within the space of a 
few minutes seven men were killed and 
twenty-two wounded, mostly from Com- 
panies I and K. Among those killed was 
2d Lt. Robert J. Bonner, commander of 
Company I. He was the fourth commander 
to have led that unit in three days and 
his death left only one officer, 2d Lt. 
Spencer M. Pitts, in the company. The 3d 
Battalion was now virtually decimated. A 
count of heads revealed that there were 
only a hundred riflemen left in it as of 
ioio. S2 

Following this debacle, Company F, 
1 06th Infantry, which previously had 
been ordered from Mount Tapotchau to 
join its parent regiment after serving for 
five days with the 2d Marine Division, 
moved down into the valley and took up 
positions in a small group of trees just 
behind Company E. S3 There the men 

Rn 27th Inf Div G-3 Jul, 28 Jun 44, Msg 22. 
Unless otherwise noted, this account of the actions 
of the 2d and 3d Battalions, 106th Infantry, is 
derived from Love, Battle for Saipan, pages 

s1 'io6th RCT Jnl, 28 Jun 44, Msg 621, In this 
message the supply vehicles are misnamed "tanks." 

S2 106th RCT Jnl, 28 Jun 44, Msg 609. 

HS Ibid., p. 99. 

were soon joined by remnants from Com- 
panies I and K as well as by the head- 
quarters of both the 2d and 3d Battalions, 
which by 13 15 had displaced forward. 

Suddenly, two Japanese tanks moved 
into view over the brow of a small hill just 
200 yards north of the crowded grove. The 
lead tank opened up on the trees with ma- 
chine guns and its 40-mm. turret gun. F,n- 
emy fire continued for ten minutes before 
a single American shot was fired in return. 
Casualties were frightful. In the 2d and 3d 
Battalions of the 106th Infantry, twelve 
were killed or mortally wounded, and 
sixty-one others were wounded. Among 
those killed were Colonel Mizony, com- 
mander of the 3d Battalion, 1st Lt. 
John T. McGregor, commanding officer 
of Company M, and Captain Tarrant of 
G Company. As the suddenness and inten- 
sity of the tank attack seemed to indicate 
a more general one, Major O'Hara, the 
senior officer present, ordered both battal- 
ions to dig in immediately. During the next 
hour a strong defensive position was con- 
structed on the ridge, and artillery fire was 
directed on the whole area to the front. 
All plans for further forward movement 
during the day were abandoned, and the 
two battalions remained where they were 
for the night. 

While this stalemate was developing in 
Death Valley proper, the 1st Battalion, 
106th Infantry, was completing its task, 
started the day before, of eliminating en- 
emy resistance in Hell's Pocket. Instead of 
holding at the south end of the area on the 
28th, Company C moved straight across 
the mouth of the pocket, while Companies 
A and B mopped up along the cliffs above. 
Company A, which was given the job of 
skirting the cliff edge on the rim of the 
pocket, encountered the most difficulty. 



Enemy mortar and rifle fire harassed the 
men throughout the morning, causing sev- 
eral casualties. Major Nichols, the battalion 
executive officer, was mortally wounded. 
By midafternoon the Japanese guns that 
were molesting the area were cleared out 
with the aid of heavy machine guns and 
mortars. Meanwhile, Company C managed 
by late afternoon to push all the way across 
the pocket and move up the left edge of 
the valley before being called back to the 
southern entrance of the valley to bivouac 
with the rest of the battalion. By nightfall 
there were still a few Japanese soldiers 
holed up in Hell's Pocket, but all organ- 
ized resistance there had finally been elim- 

At 1030 on 28 June, command of the 
27 th Infantry Division passed from Gen- 
eral Jarman to General Griner, who had 
formerly commanded the 98th Infantry Di- 
vision in Hawaii. 84 He had received his 
orders as soon as word of General Ralph 
Smith's relief had reached the headquar- 
ters of General Richardson, Commanding 
General, United States Army Forces in the 
Central Pacific Area. General Jarman had 
taken over command of the division only 
on an interim basis since he had other 
previously assigned duties as island com- 

Among the problems facing General 
Griner was the disposition of the 3d Bat- 
talion, 106th Infantry, which in its present 
state could not be expected to continue on 
the front line. Accordingly, early in the 
evening of 28 June General Griner ordered 
the 1 st Battalion, 106th Infantry, to re- 
lieve the 3d and establish contact with the 
2d Marine Division on the left.* 5 The 3d 
Battalion, 105 th Infantry, which had seen 

Maj. Gen. George W. Griner, who 
assumed command of the a; h Infantry 
Division on Saipan, 28 Jun ^944. 

no action since its withdrawal from Nafu- 
tan Point was ordered into the right of the 
division line with th.^. responsibility of 
seizing Hill Able. The 2d Battalion, 165th 
Infantry, which for the past three days had 
been attached to the 106th Infantry, was 
now attached to the 105th to assist it. S8 
With these shifts in the line ordered, the 
new division commander prepared to com- 
plete the capture of Death Valley and 
Purple Heart Ridge. 

Action on the Flanks 

In view of the success of the 4th Marine 
Division in overrunning Kagman Peninsula 
and in securing most of the O— 6 line in its 

27th Irtf Div G-3 Jnl, 28 Jun 44, Msg 32. 
1 06th RCT Jnl, 38 Jun 44, p. 106. 

s<s 27th Inf Div FO 52, 28 Jun 44, Incl to 27th 
Inf Div G-3 Jnl, 28 Jun 44, 



zone of action, Holland Smith — on the 
28th — ordered it to "hold present positions 
until further orders" and to assist the ad- 
vance of the 27 th Infantry Division by 
supporting fires from the east. 87 The 24th 
Marines and the 165th Infantry (less 2d 
Battalion), which was still attached to it, 
were ordered to establish one battalion 
apiece on the division boundary and sup- 
port the 27th Division in its movement 
along Purple Heart Ridge and Death 

On the extreme right, the 23d Marines 
held its positions on the 0-6 line and con- 
tinued to mop up the rear area and the 
caves that studded the coast line. 88 The 
24th Marines remained in division reserve 
except for the 3d Battalion, which main- 
tained its position on the boundary line 
between the 165th Infantry and the 27th 
Division. Late in the afternoon the 1st Bat- 
talion, 24th Marines, was brought up to 
fill the gap that still existed between the 
two divisions, but was unable to establish 
contact with the 27th Division on the left 
before dark. 89 

The only significant advance in the Kag- 
man Peninsula area on the 28th was in 
the zone of the 165th Infantry. There the 
3d Battalion, which was in position along 
the regimental boundary on the left flank 
of the 23d Marines, jumped off at 0630 
and within an hour had progressed to the 
O-6 line, about 440 yards forward. From 
there it moved west toward Hill 700, a 
dominating terrain feature on the division 
boundary. 1 " 1 While consolidating positions 

K7 4th Marine Div Opns Rpt Saipan, p. 28. 

KK Ibid., Annex H, 23d Marines Rpt, p. 44. 

H '' Ibid,, Annex I, 34th Marines Rpt, p. 22. 

! '° 165th RCT Jnl, aO Jun 44, Msgs 8, 2a; 4 th 
Marine Div Opns Rpt Saipan, p. 128; 165th 
RCT Rpt of Action Saipan, p. 9. 

in that area, Capt. Joseph P. Stamphcr, 
commander of Company L, was wounded 
and was replaced by ist Lt. George R. 
Weigand. 9 ' 

Meanwhile, the ist Battalion, 105th In- 
fantry, was following to the left rear of the 
3d Battalion, 165th. Moving along the tor- 
tuous mountain trails that lay in its zone, 
the battalion cleaned out the area that 
marked the boundary between the 4th Ma- 
rine Division and the 27th Infantry Di- 
vision. 92 

Earlier in the morning Colonel Kelley, 
commanding officer of the 165th Infantry, 
was struck by a fragment of a mortar shell 
and had to be evacuated. The regiment 
was turned over to Colonel Hart, its former 
executive officer, who commanded it for 
the rest of the operation, 93 

In the 2d Marine Division's zone, the 
movement toward the O-G line was again 
slowed down by the broken terrain and by 
the Japanese, who were exploiting it to 
the utmost. In the zone of the 8th Marines 
on the division right the day's objective 
was a scries of four small hills, nicknamed 
"the Pimples," across the front. 94 On the 
regimental right, the marines advanced 
with relatively little difficulty around the 
east cliffs of Mount Tapotchau before re- 
deploying on the northern slope. About 
1300 they came to a steep ravine that 
could be traversed only by a slow descent 
down a narrow crevice leading to the bot- 
tom. This took the remainder of the day. 
On the higher ground to the immediate 

1,1 165th RCT Jnl, 28 Jun 44, Msg 41. 

■> 2 Ibid., Ms S s 33, 40; 165th RCT Unit Rpt 1 1, 
28 Jun 44. 

83 165th RCT Rpt of Action Saipan, p. 10. 

" 4 This account of the action of the ad Marine 
Division is derived from: ad Marine Div SAR, 
Phase I, Forager, Sec. VI, pp. 15-16; Hoffman, 
Saipan, pp. 1 74-75- 



left, the marines made more rapid progress 
against moderate resistance and succeeded 
in pushing ahead of the units on cither 
flank. Two battalions of the 8th Marines 
were held up throughout the day by mor- 
tar and heavy machine gun fire and at 
1600, when the fighting was called off, 
were still short of the Pimples. 

In the division center, the 6th Marines 
continued to slug away at the low ridge line 
north of Tipo Pale against heavy enemy 
opposition. Medium tanks and light flame 
throwers were brought forward in an effort 
to dislodge the Japanese, but it proved to 
be an infantry-engineer task since the tanks 
could not depress their weapons sufficiently 
to reach the fortified positions that lay be- 
tween the regiment and the hills to the 
front. At the day's end the advance 
amounted only to 150 to 200 yards, al- 
though one company finally succeeded in 
wiping out the bypassed enemy pocket that 
had been occupying Tipo Pale for the past 
four days. 

South of Garapan the 2d Marines again 
held in place, fearing a forward movement 
would force them to break contact with 
the units on its right. Close air support by 
American planes accounted for the only 
casualties on the 28th suffered by the ma- 
rines in this area. In one air strike against 
Garapan, three misdirected rockets fell 
within the lines of the 1st Battalion, 2d 
Marines, causing twenty-seven casualties. 

29 June 

Action of the 2jth Division 

June 29th was a day of mixed blessings 
for the 27th Infantry Division. On the one 
hand, it was the first day since the begin- 
ning of the assault into Death Valley that 

the Army troops were able to make suf- 
ficient inroads to establish contact with the 
marines on at least one of their flanks. On 
the other hand, at the day's end, the tally 
showed another failure to seize Hill Able. 

With the 3d Battalion, 106th Infantry, 
almost depleted, General Griner had de- 
cided to replace it with the 1st Battalion, 
1 06th, which had heretofore been occupied 
with the task of cleaning out Hell's Pocket. 
Company C of the 1st Battalion was to 
guard the mouth of the pocket, but the 
rest of the battalion was ordered into the 
left of the division line with the specific 
admonition, "It is of utmost importance 
that you gain contact with the 8th Ma- 
rines today before dark." * 5 

The 2d Battalion, 106th Infantry, was 
already in position in the center of the line 
across Death Valley at the beginning of 
the day's fighting. The 3d Battalion, 105th 
Infantry, was ordered into the right to 
complete the cordon that, it was hoped, 
would squeeze the remaining life out of the 
enemy in the valley. To the right rear the 
2d Battalion, 165th, now attached to the 
105th Infantry, was ordered to finish cap- 
turing Purple Heart Ridge by overrunning 
Hill Able. 

The two advance battalions of the 106th 
Infantry jumped off on schedule but were 
at first held up by fire from what appeared 
to be dug-in tanks and machine guns lo- 
cated along S Road north of Hill Able. 98 
The 2d Battalion requested artillery fire 
but was denied it because the area in 
question was too close to the boundary line 
of the 2d Battalion, 165th Infantry, the 

,JS 1 06th RCT Jnl, 29 Jun 44, Msg 679; 27th 
Inf Div FO 52, 28 Jun 44, Incl to 27th Inf Div 
G-3 Jnl, 28 Jun 44. 

96 27th Inf Div G-3 Jnl, 29 Jun 44, Msgs 17, 

27, 28, 34. 



battalion that was supposed to be working 
up Hill Able from the south. 97 

To add to the delay, friendly artillery 
fire began to fall on the front lines of the 
3d Battalion, 105th Infantry, and it had 
to reorganize before it could jump off. 98 
Finally, Company K got lost and was in- 
volved in an extracurricular fire fight be- 
fore reaching the right flank of the front 
line in Death Valley to which it was as- 
signed. 88 

Company K had moved out of the 
battalion assembly area around 0700 and 
had skirted the inside (west) slope of 
Purple Heart Ridge until it reached Hill 
King. There, one of the guides furnished 
by the 106th Infantry informed Capt. 
Alexander C. Bouchard, the company com- 
mander, that Hill King was cleared and 
offered an easy route to the battalion line of 
departure. Unfortunately, Hill King had 
been lost to the enemy two days before 
and had not been recovered. As K Com- 
pany began to climb to the top it came 
under severe enemy rifle fire, which was 
soon followed by an American artillery 
barrage designed to clear the hill of Japa- 
nese troops in advance of the attack of 
Company E, 165th Infantry, which was 
coming up from a different direction. 
Nineteen of Bouchard's men were immedi- 
ately wounded, and the rest scattered. 
Company E then assaulted the hill and, 
with some help from K Company, cleared 
the remaining Japanese, most of them hid- 
den in foxholes. 100 This eliminated Hill 
King as a source of trouble, but Company 
K had to stop and reorganize before ad- 
vancing to its line of departure. Not until 

87 106th RCT Jnl, 29 Jun 44, Msg 700. 
98 27th Inf Div G-3 Jul, 29 Jun 44, Msgs 12, 
13 ; 105th RCT Jnl, 29 Jun 44, Msg 2. 
"» 105th RCT Jnl, 29 Jun 44, Msgs 5, 6. 
100 Love, Battle for Saipan, pp. 473-74- 

about 1 300 was it able to take position on 
the right of Company I and close the line 
across Death Valley. 101 

Finally, about 1400, all three battalions 
— the 1 st and qd of the 106th, and the 3d 
of the 105th — jumped off in line abreast 
toward the northern end of Death Val- 
ley. 102 Within less than twenty minutes the 
106th Infantry had advanced 400 yards; 
the 105th was only a little behind. By 1445 
another 300 yards had been gained. By the 
end of the day the 1 06th had scored a total 
gain of 900-1,000 yards, while the 3d Bat- 
talion, 105th, had moved ahead about 600 
yards.' 03 By 1530 the men of 1st Battalion, 
1 06th, had visual contact with the 2d Ma- 
rine Division on the northern slopes of 
Mount Tapotchau on their left. 101 

At last the long-broken link was restored. 
At last General Holland Smith, corps 
commander, could find good words to say 
for the 27th Division. Viewing its progress 
through Death Valley on the 29th from 
the vantage point of Mount Tapotchau, he 
"expressly complimented" the division's 
performance to General Griner, its new 
commander. 105 

No such good fortune attended the ef- 
forts of the 2d Battalion, 165th Infantry, 
on the division right flank. After recaptur- 
ing Hill King, Company E moved off at 
T335 toward Hill Able. 106 A shower of 
hand grenades greeted the men who tried 
to scale the steep southern slope. Com- 

1U1 105th RCT Jnl, 29 Jun 44, Msgs 13, 14; 
iof,lh RCT S-3 Dispatch Summary, 29 Jun 44, 
Incl to 27th Inf Div G-3 Jnl, 29 Jun 44. 

102 27th Inf Div G-3 Jul, 29 Jun 44, Msg 50; 
105th RCT Jnl, 29 Jun 44, Msg 14 10. 

103 27th Inf Div G-3 Jnl, 29 Jun 44, Msg 5,0; 
ioGth RCT Jnl, 29 Jun 44, Msgs 729, 733; 27th 
Inf Div G-3 Periodic Rpt 13. 

104 2 v, tn j n f d; v q_2 jni^ 29 Jun 44, Msg 59. 

105 27th Inf Div Fid Msg 2, 29 Jun 44, Incl to 
27th Inf Div G-3 Jnl, 29 Jun 44. 

'"« .05th RCT Jnl, 1335, =9 J™ 44- 




Truck- Mo un ted Rocket Launchers Firing at hills north of 8th Marines 
position on sg June. 

pany E withdrew. Meanwhile, Company 
F on the right was making slow progress in 
a move to come up Hill Able from the cast. 
Continuous mortar fire from the hill 
impeded its movement, and one shell 
wounded Lieutenant Trammel, the com- 
pany commander. By late afternoon the 
company had worked its way into a posi- 
tion just east of Hill Able, but by that 
time the hour was too late to warrant an 
attack. 1 " 7 

Action on the Flanks 

The 4th Marine Division again spent the 
day patrolling and consolidating its posi- 

107 Love, Battle for Saipan, pp. 480-85. 

tions while waiting for the 27th Infantry 
to reach the 0-6" line and establish contact 
on its left flank. 10K The 23d Marines sent 
out patrols as far as 1,200 yards to the 
front, capturing small groups of unarmed 
Japanese. 10tf To the left rear of the divi- 
sion zone, elements of the 24th Marines 
received fairly heavy machine gun and 
mortar fire from enemy groups that had 
apparently filtered in to escape the pressure 
of the 27th Division. 110 The 3d Battalion, 
165th Infantry, continued to improve its 
positions by consolidating on Hill 700 on 
the left flank of the division zone and on 

108 4th Marine Div OprlS Rpt Saipan, p. ag. 
103 Ibid., Annex H, 23d Marines Rpt, p. 44. 
110 Ibid., Annex I, 24th Marines Rpt, p. 32. 



Charan Danshii ridge to the north. In so 
doing, the battalion fell under consider- 
able artillery and mortar fire, from enemy 
positions on its left front in the zone of 
the 27th Division. Around 1400 the fire 
reached such intensity that the battalion 
withdrew from Charan Danshii and dug in 
for the night on Hill 700. ni Colonel Hart 
informed the Marine division headquarters 
that in order to hold Hill 700 he would 
have to move the 3d Battalion about 300 
yards to the west, leaving about the same, 
distance between his troops and the 23d 
Marines. To fill this gap, some eighty men 
of the 1 st Battalion, 165th, were sent into 
the front lines, their own previous positions 
being taken over by elements of the 3d 
Battalion, 24th Marines. 112 

On the corps left the 2d Marine Division 
again made slow progress against rugged 
terrain and determined enemy resistance. 
On the division right the 8th Marines 
struggled through dense jungle until about 
noon, when it was Stopped by heavy fire, 
from the Pimples to the north. Two well- 
placed rocket concentrations were put on 
these hills with excellent coverage of the 
area, but before the regiment could move 
forward sufficiently to undertake its as- 
sault it was too late to justify an attack, 
and the hills remained uncaptured for an- 
other day. 113 

In the center, a readjustment of the 
lines of the 6th Marines took so long that 
the attack did not jump off until about 
noon and made little progress during the 
rest of the day. Once again, too, the 2d 

111 165th RCT Unit Rpt 12, 29 Jun 44; 165th 
RGT Jnl, 29 Jun 44, Msgs 37, 39. 

112 165th RCT Jnl, 29 Jun 44, Msgs 28, 29. 
li ' J This account of the action of the 2d Marine 

Division is from: 2d Marine Div SAR, Phase I, 
Forager, Sec. VI, p. 18; Hoffman, Saipan, pp. 

Marines below Garapan were forced to 
wait for the rest of the division to come 
into line. Their day was not spent in idle- 
ness, however. A group of Japanese of 
about platoon size had dug in on a small 
hill some 500 yards ahead of the Marine 
lines in such a manner as to defy extermi- 
nation by artillery or mortars. In order to 
entice the enemy out of his underground 
caves and passageways, the 2d Marines 
simulated an infantry attack on the morn- 
ing of 29 June. Following a heavy artillery 
preparation, the marines opened with small 
arms and machine guns as though prepar- 
ing for an assault, then ceased their ma- 
chine gun fire but continued with small 
arms to heighten the illusion of an infan- 
try attack. At this the Japanese emerged 
to man their machine guns and auto- 
matic weapons, whereupon they were im- 
mediately wiped out by American artillery 
and mortars. Thus, the way was cleared 
for a relatively easy entry into Garapan 
once the 2d Division's lines had been suffi- 
ciently straightened to justify the move- 

30 June 

Death Valley: Capture 
and Breakthrough 

The last day of June witnessed the end 
of the long and bitter struggle of the 27th 
Division to capture Death Valley and 
Purple Heart Ridge. As General Schmidt, 
USMC, later testified, "No one had any 
tougher job to do." 114 General Griner's 
orders for the day read that operations to 
reduce Hill Able would "be concluded" by 
the 2d Battalion, 165th Infantry. 115 The 

111 Ltr, Gen Schmidt, USMC (Ret.), to Gen A. 
G. Smith, 10 Jan 55 (Incl), OGMH. 

11 5 27th Inf Div Fid Msg 2, 29 Jun 44, Incl to 
27th Inf Div G-3 Jnl. 



orders were complied with. Jumping off at 
0715, the battalion launched its attack, 
with Company F assaulting the hill from 
the east (right), G from the west. Com- 
pany E remained back on Hill King to 
lend fire support. 116 Opposition was light 
and by 0940 the hill was reported se- 
cured. For the rest of the day the battalion 
dug in and consolidated its positions, 117 

Meanwhile, the troops in the valley be- 
low sustained the momentum of the 
previous day. Since Colonel Bradt's 3d 
Battalion, 105th Infantry, had lagged be- 
hind the 106th on the 29th, the main effort 
was to be in his zone, on the right of the 
division line. 

Bradt's battalion, accompanied by nine 
tanks, jumped off promptly at 0715 fol- 
lowing a fifteen-minute artillery prepara- 
tion. 11 ** After about two hours of fairly 
easy going the battalion came under fire 
from Hill Uncle- Victor, located about 
1,400 yards north of Hill Able. "It ap- 
pears," reported Bradt, "to be another 
Hill Able." 11H Colonel Stcbbins then pro- 
posed to place an artillery barrage in the 
area, but could not safely do so until the 
tanks that were operating at the foot of 
the hill withdrew, so the attack on the 
division right came to a temporary stand- 
still. Around noon the artillery preparation 
was completed, and Colonel Bradt's troops 
moved up onto the troublesome hill and 
declared it secured, 120 

Meanwhile, at two points on the division 
boundary contact was at last established 
between the 37th Infantry Division and 
the 4th Marine Division. Soon after 1000 
a patrol sent out from the 3d Battalion, 

1111 27th Inf Div G-3 Jnl, 30 Jun 44, Msgs 7, 15. 
117 Ibid., Msgs a 1, 24. 
llB Ibid., Msg 7. 
1 1 <* Ibid., Msg 1 7. 
120 Ibid., Msg 46. 

105th, made contact with a party of ma- 
rines at a point about 600 yards northeast 
of Hill Able. 121 Two hours later a platoon 
of the 27th Reconnaissance Troop that had 
in the morning been fed into the right of 
the division line for that express purpose, 
established contact with the 1st Battal- 
ion, 24th Marines, just northeast of Hill 
Able. 122 At long last the 4th Marine Di- 
vision was in a position to move forward 
without fear of exposure on its left flank. 

In the center of the 27th Division's line, 
ioGth's ad Battalion made fairly steady 
progress throughout the day against light 
opposition, but on the left its 1st Battalion 
discovered to its sorrow that the cliffs of 
Mount Tapotchau were still not entirely 
cleared of the enemy. Mortar and machine 
gun fire sporadically harassed the battal- 
ion as it tried to keep pace with the rest 
of the division in its move to and through 
the northern end of Death Valley. 

Late in the afternoon, 2d Lt. Ralph W. 
Hill of the 1 st Platoon, Company B, lo- 
cated at least one of the positions in the 
cliffs that had been causing the battalion 
so much trouble. Taking two enlisted 
men with him, and against the advice of 
his company commander, he reconnoitered 
the cliff line and discovered a machine gun 
position on a tiny ledge about thirty feet 
up the side of the cliff. The party was soon 
detected by the Japanese. Hill was shot 
and then all three Americans were 
wounded by the explosion of a hand gren- 
ade. The two enlisted men retreated but 
came back later to find their platoon 
leader, undaunted by his serious wounds, 
still firing at the enemy position, which was 
forthwith silenced. Whether this action 
was decisive or not is unknown, but by the 

iai Ibid., Msg 41. 
12:1 Ibid., Msg 54. 



next morning the cliff was clear ol Japa- 
nese, and it was assumed that they had 
come down from the caves and withdrawn 
to the north during the night. 12;J 

All together, the advance of the 27th Di- 
vision's line on 30 June was about 400 
yards. 124 By the day's end, physical con- 
tact had been established on both the right 
and the left with the two Marine divi- 
sions, 125 Death Valley had been left be- 

Action on the Flanks 

Once more the 4th Marine Division on 
Kagman Peninsula spent the day resting, 
patrolling, and consolidating its lines. Along 
the coast the 23d Marines pushed its pa- 
trols as deep as 800 yards north of the 
0-6 line but, aside from capturing a few 
civilians, made no contact with the enemy. 
With the advance of the 27th Division, the 
1st Battalion, 24th Marines, which was the 
southernmost Marine unit disposed along 
the division boundary line, was pinched 
out. Throughout the day little enemy ar- 
tillery and mortar fire was received, al- 
though small arms fire from pockets of re- 
sistance located in the 27th Division's /one 
continued to cause some casualties. 12 '' 

On the left of the 27th Division, the 2d 
Marine Division continued its slow prog- 
ress through the wooded hills and ravines 
north of Tapotchau. The 2d Battalion, 8th 
Marines, captured one of the Pimples, and 
a medium tank platoon from the Army's 
762d Tank Battalion came up in the late 

afternoon to lay fire on the remaining hills 
similarly nicknamed. In the division center, 
the 6th Marines gained little ground but 
was able to straighten its lines, while on 
the left the 2d Marines again waited on 
the outskirts of Garapan for the rest of 
the corps line to come abreast. 127 

Central Saipan: Sum-up 

With the closing of the gaps on either 
side of the 27th Division's line, the battle 
for central Saipan can be said to have 
come to a successful end. The cost had 
been high and the progress painfully slow. 

Total American casualties came to an es- 
timated 3,987. 1S8 Of these, the 4th Marine 
Division suffered 1,506; the 2d Marine Di- 
vision, 1,016; the 27th Infantry Division, 
1,465. The Army division was especially 
hard hit among its line officers. The 165th 
Infantry lost its commander, Colonel 
Kelley, who was wounded in action. 
Colonel McDonough, commander of the 
2d Battalion, 165th, was wounded and 
evacuated; his successor, Major Brusscau 
was wounded and later died; and Colonel 
Mizony of the 3d Battalion, 106th, was 
killed in action. In addition, a total of 
twenty-two company commanders of the 
165th and 1 06th Regiments were either 
killed or wounded in action during the 

12:j This action is described in Love, Baltic For 
Saipan, pages 606-08. 

121 27th Inf Div G-3 Jnl, 30 Jun 44, Msg 94; 
Love, Battle for Saipan, p. 600. 

135 27th Inf Div G-3 Jnl, 30 Jun 44, Msgs 32, 

4 I .. r >4- 

120 4th Marine Div Opns Rpt Saipan, pp, 29-30. 

137 2d Marine Div SAR, Phase I, Forager, pp. 
19-20; Hoffman, Saipan, pp. 183-84. 
~ J2 « NTLF Rpt Marianas, Phase I, Incl C, G-i 
Rpt, App. 1 ; 27th Inf Div G-i Periodic Rpts 7 
through 14. Casualty figures for the 27th Division 
are derived from the second source cited, since 
NTLF figures for the Army division arc obviously 
incomplete. In computing 27th Division casualties 
for central Saipan for the period 33-26 June, the 
figures for the 105th Infantry have been deducted 
from the total division casualties, since that regi- 
ment was either in reserve or fighting on Nafutan 
Point during the period. 



In the center of the corps line, it had 
taken the 27th Division eight days to ad- 
vance 3,000 yards. In the same time the 
2d Marine Division had advanced 2,600 
yards on its right and 1,500 yards on its 
left where any further forward movement 
was unfeasible until the rest of the line 
came abreast. Only on the corps right, in 
the zone of the 4th Marine Division, had 
the troops advanced rapidly. By the close of 
30 June this division, with its Army at- 
tachments, had pushed about 4,400 yards 
cast to the tip of Kagman Peninsula and 
about 5,000 yards northwest from its orig- 
inal line of departure of June 23d. Un- 
doubtedly, it would have gone farther had 
it not been held back by the relatively slow 
advance in the center of the corps line. 

This bald account of the yardage gained 
is by no means a true measure cither of the 
difficulties of the fighting or of the results 

achieved. Unlike the 4th Marine Division, 
the other two divisions faced extremely dif- 
ficult terrain, which the Japanese, in spite 
of their dwindling strength, exploited to 
the utmost. The main drive of the 27th 
was up the long axis of a valley flanked 
on both sides by fortified hills, cliffs, and 
mountains. That of the 2d Marine Division 
was across the largest and most precipitous 
mountain mass on the island. Against any 
but a completely prostrated enemy, the as- 
sault could only have been slow and 

Whatever the cost of the drive, the re- 
sults were decisive. Mount Tipo Pale and 
Mount Tapotchau were captured; Death 
Valley, Purple Heart Ridge, and Kagman 
Peninsula occupied. The main line of re- 
sistance set up by the Japanese after 
their withdrawal from the beachhead was 
broached and overrun. 


The Capture of Northern Saipan 

Drive to Tanapag 

1-2 July 

With Death Valley cleared, Holland 
Smith was at last in a position to push his 
forces rapidly ahead and seal off the Jap- 
anese remaining in the northern neck of 
Saipan. On i July he established the next 
corps objective at a line (O-7) that cut 
across the base of this neck in an arc, at 
a distance from about 1,000 yards on the 
right to 6,000 yards on the left from the 
respective flanks of the corps front. Be- 
tween line O-7 and the corps front, as of 
1 July, lay the hill mass (Hills 221 and 
112 meters) 1 on which the Japanese had 
chosen to anchor their last defensive posi- 
tion across the island. The attack was to 
be made with the three divisions abreast in 
the same order as before, the main effort 
again to be in the center, i n the zone of the 
27th Division ? \{Map //)| 

In the earlier phases of the fighting on 
Saipan, General Holland Smith had noted 
a tendency on the part of his infantry com- 
manders to neglect the abundant artillery 
support available to them, and to rely too 
heavily on their own weapons. Too fre- 
quently, he believed, the front-line troops 

1 American troops called Hill 221 "Radar Hill." 
On the American map, Hill 112 was located just 
southeast of Tarahoho. 

2 NTLF Opn Order 19-44, ! J u ' 44- 

had failed to call for massed artillery fires 
before jumping off in attack. Moreover, 
even when artillery concentrations had 
been properly called for, they were often 
not followed promptly by tanks and infan- 
try, and thus the whole effect of the artil- 
lery preparation was wasted/ To correct 
this situation the corps commander specif- 
ically ordered : 

Massed artillery fires will be employed to 
support infantry attacks whenever practic- 
able. Infantry will closely follow artillery 
concentrations and attack ruthlessly when 
the artillery lifts. Absence of tanks is no ex- 
cuse for failure of infantry to press home the 
attack. 4 

Even before the order calling for a quick 
thrust to Tanapag had been issued, the 
corps line had been pushed forward and 
straightened in preparation for the drive. 
In the center the 27th Division, on 1 July, 
registered a gain of about 400 yards on 
the right and 600 on the left against mod- 
erate opposition. 5 On the right, the 4th 
Marine Division maintained its positions 
and sent patrols as far out as 1,500 yards 
in front of its line without establishing con- 
tact with the enemy/' It was clear that the 

3 NTLF Dispatch 01 1806 to 2d Marine Div, 4th 
Marine Div, 27th Inf Div, Incl to 27th Inf Div 
G-3 Jnl. 

4 NTLF Opn Order 19-44, 1 Jul 44. 

" J y7th Inf Div G-3 Periodic Rpt 15, I Jul 44. 

6 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RCT Rpt, p. 45; Ibid., Annex I, 24th RCT 
Rpt, p. 62. 

Mantogatso I 

Tanapag coconut 


Floras Pi 



* «■••*•••• 


Mt Tapolchau 

MAP 11 






1-4 July 1944 

»»)i»»)»»»))> FftONT LINE, fvENlNS JO JUNE 

oooooooo Positions reached I July 

!^3t3c= Positions reached i M\.t 

• ••••••• Positions beached 3 Jul* 

iiiiiiiiiiiiih positions reached 4 july 

Contour inter ml 100 felt 

O 900 

| h > | I I 

(000 YARDS 

1000 METERS 

F Twmpt» 



Japanese were retreating to the north. 
Early on i July members of the 27th Di- 
vision had seen a small body of Japanese 
lugging ammunition up one of the roads 
that led out of Death Valley. All morning 
long, a 4th Marine Division observation 
post atop Hill 700 reported, the Japanese 
had been retreating in groups of three or 
four, carrying their packs and equipment 
with them. 7 

While these events were taking place on 
the right and in the center of the line, the 
2d Marine Division gained more yardage 
than on any other day since the landing. 
The strong line of resistance through 
Mount Tapotchau had been smashed. Over 
terrain that was far better suited to the 
employment of tanks than the cliffs and 
defiles of Tapotchau, the 8th Marines ad- 
vanced 800 yards. On its left, the 6th Ma- 
rines kept pace in spite of having to over- 
come several pockets of heavy resistance, 
and on the extreme left the 2d Marines 
continued to patrol south of Garapan in 
preparation for the long-awaited push into 
the city itself, which was scheduled for 2 
July. 8 

On 2 July, the 4th Marine Division, 
which had spent most of its time dur- 
ing the past days resting and patrolling, 
plunged ahead for about 1,500 yards in its 
zone. Resistance was so light that the as- 
sault battalion of the 24th Marines suffered 
only one man wounded during the day." 

On the 4th Division's immediate left, the 
3d Battalion, 165th Infantry, which had 
now been returned to the control of its 

7 NTLF Rpt Marianas, Phase I, Incl D, G-2 

Rpt, PP- 43-44- 

8 ad Marine Div SAR, Phase I, Foracjkr, Sec. 
VI, pp. 20-21. 

8 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RCT Rpt, p. 45; Ibid., Annex I, 24th RGT 
Rpt, p, 22; Hoffman, Saipan, p. 193. 

parent division, maintained the same pace. 
By 1400 Major Focry's men had pushed 
ahead about 1,700 yards, 10 leaving behind 
at about the same distance to their left 
rear the 3d Battalion, 105th Infantry. This 
unit was held back by intense rifle and 
machine gun fire, leaving a deep re-entrant 
between the 3d Battalion, 165th, on its 
right and the 106th Infantry on its left. 
The latter had succeeded in advancing 
about i,ooo yards after clearing out five 
enemy tanks emplaced as pillboxes. To 
close the gap, General Griner late in the 
afternoon ordered the 1st Battalion, 105th 
Infantry, which had been in regimental re- 
serve, to make a wide end run around the 
regiment's left flank, bypass the enemy 
strongpoint that was holding up the 3d 
Battalion, 105th, and establish contact 
with the left flank of the 3d Battalion, 
165th. 11 This the 1st Battalion did by 
1800. 12 

In the zone of the 2d Marine Division, 
the two regiments on the right made good 
progress (800-1,200 yards) during the day 
in spite of rough terrain and the fact that 
the 8th Marines was temporarily disor- 
ganized when friendly artillery fire fell into 
its lines causing forty-five casualties. On 
the division left flank the 2d Marines, after 
its prolonged wait before Garapan, was at 
last ordered to enter the city. As it did so, 
the devastating effect of the many days of 
artillery bombardment and naval shelling 
was revealed on all sides. Garapan was 
little more than a mass of rubble, and 
though there was some hostile rifle fire, the 
2d Marines quickly occupied the center of 

10 165th RCT Jnl, 2 Jul 44, Msg 58. 

11 27th Inf Div G-'i Jnl, 2 Jul 44, Msg 62. 

12 105th RCT Opns Rpt Forager, p. 8; 106th 
RCT Opns Rpt Forager, p. r8; 27th Inf Div G-3 
Periodic Rpt 16, 2 Jul 44. 



Marines Dash Through Blazing Garapan, 2 July. 

the town with the help of tanks and ar- 
mored amphibians. To the immediate 
cast the Japanese, entrenched on a hill 
overlooking the city, caused considerable 
trouble, but by nightfall enemy resistance 
had subsided, and the regiment dug in 
about 700 yards from its morning line of 
departure. 13 

Under the mounting pressure of the 
American attack, the Japanese on the night 
of 2 July once more fell back to new po- 
sitions. Six days earlier General Saito had 
decided to make his last stand along a line 
running from north of Garapan through 
Radar Hill and Hill 112 f meters ) on to 

the coast. 14 Now those troops able to do 
so were to retire to the new line. It was 
high time. Many of them had been so 
pressed for provisions that they were eating 
field grass and tree bark. 15 

3-4 J u b 

The axis of the drive to Tanapag Harbor 
now took a more northwesterly direction, 
with the main effort still in the center in 
the zone of the 27th Division. The Japa- 
nese were retreating rapidly and in a piece- 
meal manner. Saito's plans for an orderly 
withdrawal to the north were obviously 

3 3 2d Marine Div SAR, Phase I, Forager, 
VI, p. 22; Hoffman, Saipan, pp. 186-87. 


1 * Japanese Studies in World War II, 55. 
x *'lbid. 



Infantrymen Observe Hill 767 before approaching. Note abandoned Japanese 
37-mm. gun in foreground. 

breaking down in the face of the gathering 
momentum of the attacking troops. 

On the morning of 3 July the attack on 
the right got off to a slow start as a result 
of confusion shared by the 4th Marine Di- 
vision and the 1st Battalion, 165th, as to 
the intentions of each. The Army unit was 
prepared to jump off on schedule at 0800, 
but held back because an air strike in front 
of the Marine division's lines prohibited 
forward movement. After the strike the 
soldiers continued to hold, waiting for the 
marines to go forward. The latter made no 
move on the false assumption that the 1st 
Battalion, 165th, was waiting for the unit 
on its left to come abreast. This misunder- 

standing continued until 1100, when the 
Marine division and the Army battalion 
jumped off together. 16 

The 4th Marine Division attacked in 
columns of battalions, with the 3d Battal- 
ion, 23d Marines, 3d Battalion, 25th Ma- 
rines, and 1 st Battalion, 24th Marines, in 
the assault from right to left. After a few 
hours of fairly unimpeded movement, the 
battalion on the right was pinned down by 
heavy machine gun, mortar, and rifle fire 
from well-concealed positions in caves and 
wooded recesses on Hill 721 and on a nose 

16 165th RCT Record of Opns Forager, Sec. 
IV, S-3 Rpt, p. 7; 27th Inf Div G-3 Jnl, 3 Jul 
44, Msgs 1 1 123, 1 1 127; Hoffman, Saipan, p. 202. 



abutting south from it that the marines 
were later to dub "Fourth of July Hill." 
Several attempts were made to penetrate 
the position by both frontal assault and 
envelopment, but each time the troops were 
so badly shot up that they were forced 
to retire. The approaches were heavily 
mined, and neither tanks nor self-propelled 
mounts could come in close. Finally, at 
1 7 15, after three hours of continuous 
fighting, the battalion pulled back 300 
yards to safe positions and let the artillery 
take over. All night long howitzers of the 
14th Marines pounded the strongpoint and 
kept it neutralized. 17 

Daylight of 4 July revealed that the Jap- 
anese had either withdrawn during the 
night or been eliminated by the intense ar- 
tillery fire, and by 1 1 35 both Fourth of 
July Hill and Hill 721 were in American 
hands. Within another hour a battalion of 
the 23d Marines had moved 800 yards to 
the northeast and had taken Hill 767 with- 
out opposition. 18 Meanwhile, the other 
two regiments of the 4th Marine Division 
had kept abreast. The 25th tied in with 
the 23d around Hill 767 on the night of 4 
July, by which time most of the 24th Ma- 
rines had been relieved by the 165th In- 
fantry. 19 

In the zone of the 27th Division, the 3d 
Battalion of the 165th kept abreast of the 
marines on 3 July without meeting more 
than sporadic fire from the Japanese. 20 To 
its left the 1st Battalion, 105th, moved 
even faster against negligible opposition 

17 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RCT Rpt, pp. 45-46. 

1B Ibid., v . 46. 

10 4th Marine Div Opns Rpt Saipan, Annex I, 
24th RCT Rpt, p. 23; Ibid., Annex J, 25th RCT 
Rpt, pp. 7-8 ; Hoffman, Saipan, pp. 203-06. 

ao 165th RCT Record of Opns Forager, Sec. 
IV, S-3 Rpt, p. 7. 

and by 14 10 reached its objective for the 
day — the high ground 2,000 yards east of 
Tanapag Harbor overlooking the plains of 
Tanapag. 21 

The 106th Infantry on the division left 
jumped off on schedule and also reached 
the high ground north of Tanapag by late 
afternoon. Earlier in the morning the 1st 
Battalion had found a pocket of Japanese 
close to the division boundary line, but 
these men were quickly silenced by tanks 
and self-propelled mounts. Thereafter op- 
position was light. Meanwhile, to the rear 
of the front line the 3d Battalion was mop- 
ping up a bypassed enemy position in the 
cliffs north of Tapotchau. While one com- 
pany of the 8th Marines tried to get at 
the Japanese-infested caves from above, 
Company K, 106th, contained the enemy 
from the plain below. After this maneuver 
failed to produce results, artillery was called 
upon to lay down a concentration. This too 
accomplished nothing, and at nightfall the 
strongpoint was still in enemy hands. 22 

In the early morning hours of 4 July a 
large group of Japanese, trying to escape 
to the north to join General Saito, stumbled 
into the command post of the 165th In- 
fantry. After a brisk fire fight, twenty- 
seven of the enemy were killed including 
a number of officers, one of whom proved 
to be Colonel Ogawa, commanding officer 
of the 136th Infantry Regiment. 2 * 

On his body Ogawa carried Saito's with- 
drawal order of 2 July. Ogawa himself had 
ordered the remnants of his own regiment, 
now bypassed by the Americans, to com- 
mence their withdrawal at 2200 on the 

21 105th RCT Opns Rpt Forager, pp. 8-9. 

22 1 06th RCT Opns Rpt Forager, Sec. I, pp. 

as 165th RCT Record of Opns Foracer, Sec. 
IV, G-3 Rpt, p. 7; NTLF Rpt Marianas, Phase 
I, Incl D, G-2 Rpt, p. 49. 



night of the 3d. When he was killed, 
Ogawa was bound for his new command 
post, which he hoped to locate on a cliff 
about 500 meters cast of Hill 22 1. 24 
Ogawa was not merely in command of a 
decimated regiment, but of the entire Jap- 
anese left and thus one of the few key men 
remaining among the Japanese defenders. 
His death was a heavy blow to the already 
stunned and reeling enemy, but the circum- 
stances of his death indicate that an even 
greater misfortune had befallen the Japa- 
nese. It is more than likely that many if 
not most of the units under Ogawa' s 
command behind the American lines never 
reached their assigned positions to the 
north. Thus the Japanese left flank, toward 
which the main drive of the American 
forces was now oriented, lay weakened, ex- 
posed, and almost leaderless. 

The Fourth of July was to see the culmi- 
nation of the 27th Division's thrust to 
Tanapag Plain. On the right the 1st Bat- 
talion, 165th, jumped off at 0730 on 
schedule and, meeting almost no opposi- 
tion, quickly pushed forward to the last 
low ridge line overlooking the Flores Point 
seaplane base. A heavy downpour, the first 
daylight rainfall of any severity since the 
landing on Saipan, mired the tanks, but 
it made little difference since there were no 
targets at which they could fire. The rest 
of the regiment failed to keep pace so, from 
1030 until midafternoon when new orders 
were issued changing the direction of the 
attack, the men of the 1st Battalion rested 
atop the ridge and took pot shots at the 

Japanese milling in the coastal valley be- 
low. 25 

In the center of the division line, the 1st 
Battalion, 105th, made rapid progress to a 
position just beyond the same ridge line, 
where it found a strongpoint manned by 
about three hundred enemy soldiers, with 
some machine guns. A called artillery bar- 
rage scattered the Japanese, and by 1600 
most of the battalion had succeeded in 
reaching the beach. 26 On the left, the two 
assault battalions of the 106th had an eas- 
ier time in spite of the heavy" undergrowth 
through which they had to push. By 1430 
they made their way into the Flores Point 
seaplane base, where they were joined in 
mopping-up operations by the 8th Marines, 
To their rear the 3d Battalion, 106th, spent 
most of the day finishing off the trouble- 
some caves that had occupied it on the 3d, 
First, flame thrower teams went forward 
to destroy the ring of enemy machine gun 
positions that had been protecting the 
largest cave. Next, a public address system 
was brought up and interpreters broadcast 
pleas to the main body of Japanese to sur- 
render. When this failed, the infantry re- 
sumed the attack and reduced the position. 
It yielded one wounded Japanese and fifty 
dead plus an unknown number sealed up 
in the smaller caves adjacent to the main 
position. Tt was the last Japanese strong- 
point remaining on the slopes below Ta- 
potchau. 27 

In the zone of the 2d Marine Division, 
by nightfall of 3 July the 6th Marines was 
still held up on the ridge line about 1,000 

U4 In Japanese terrain designation, this was Hill 
205 (meters). NTLF Rpt Marianas, Phase I, Incl 
D, G-2 Rpt, map, pp. 50-51; CINCPAC- 
GINCPOA Trans 10531, excerpts from a notebook 
of field orders, 14 Jun-3 Jul. 

a ' 165th RCT Record of Opns Forager, Sec. 
IV, G-3 Rpt, p. 7 ; Love, Battle for Saipan pp. 
670-72, 680. 

2,f 105th RCT Opns Rpt Forager, p. 9; 27th 
Inf Div G-3 Periodic Rpt 18. 

a7 1 06th RCT Opns Rpt Forager, p. 20. 



Troops Searching Out the Enemy in the Tapotchau cliff area on 4. July. 

yards from the ocean shore, but the 2d 
Marines had finished mopping up Garapan 
and had pocketed the small enemy garrison 
remaining on the tip of Mutcho Point. 
Next day, both regiments reached the shore 
line. 28 

During most of the day the Japanese, 
under relentless pressure from the attack- 
ers, had been retreating steadily toward 
Saito's last headquarters, the rallying point 
for the final desperate counterattack that 
would come two days later. The Japanese 
commander had set up his command post 
in the valley running south from the village 
of Makunsha — appropriately enough la- 

beled "Paradise Valley" by the Americans 
and "Hell Valley" by the Japanese?" A 
captured Japanese officer was later to de- 
scribe in moving terms the miserable situa- 
tion in which Saito and his staff found 

This area is generally called the Valley of 
Hell and we felt that this was an unpleasant 
hint and suggestion concerning our future. 

The intelligence which managed to reach 
me at this last place was all depressing. 

On 4 July, an enemy unit appeared on the 
other side of the valley and fired at us with 
heavy automatic weapons. At that time I felt 
wc were entirely surrounded and had lost all 

28 2d Marine Div SAR, Phase I, Forager, Sec. 
VI, p. 23; Hoffman, Saipan, pp. 198-207. 

29 NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, pp. 51-52. 



General Saito was feeling very poorly be- 
cause for several days he had neither eaten 
nor slept well and was overstrained. He was 
wearing a long beard and was a pitiful sight. 

That morning that very valley received in- 
tense bombardment (I don't know whether 
it was naval gunfire or pursuing fire from 
artillery, but it was the seeond most intense 
bombardment I had been in) . It was so fierce 
that I thought maybe the cave where the 
headquarters was would be buried. At this 
time the staff and Lt. Gen Saito received 
shrapnel wounds. 

I felt that the final hour was drawing 
near. 30 

Change of Direction 

As it became apparent that the drive to 
Tanapag Harbor could be successfully con- 
cluded on the 4th, General Holland Smith 
prepared plans for the last phase of the 
Saipan campaign. The direction of the 
drive would change to the northeast — 
toward Marpi Point and the remaining 
Japanese airfield, which bore the same 
name. Most of the 2d Marine Division, 
which had by now been pinched out, was 
assigned to corps reserve. The final assault 
was to be conducted with the 4th Marine 
Division on the right, 27th Infantry Di- 
vision on the left. To allow time for the 
necessary shifts, jump-off hour was set at 
noon, 5 July. 31 General Grincr was or- 
dered to relieve the two left battalions of 
the 4th Marine Division. The division 
boundary line would now cut down the 
northern end of the island slightly west of 
the middle. Griner decided to commit the 
165th Infantry on the right, the 105th on 
the left, the 106th going into reserve. 32 

Late in the afternoon he ordered the 2d 
and 3d Battalions, 165th, to relieve the 
marines in that portion of the line now as- 
signed to the Army division. In effecting 
this relief, contact with the 1st Battalion, 
165th, on the regimental left was lost, 
so Griner ordered Colonel Bradt's 3d Bat- 
talion, 105th, to fill the gap. Unfortun- 
ately, Bradt's orders were garbled in trans- 
mission and he moved to the left rather 
than to the right of the 1st Battalion, 
165th. By this time night had fallen, and 
before the error could be rectified almost 
a hundred Japanese were able to infiltrate 
through the gap and harass the front-line 
troops throughout the night. The attacks 
were sporadic, however, and by morning 
the gap had been filled , and the en emy re- 
pulsed or destroyed. 3S \(Map III.)\ 

5 My 

4th Marine Division 

On the right half of the corps line, Gen- 
eral Schmidt placed the 25th Marines on 
the right and the 24th on the left, and or- 
dered the 23d to clean up the area between 
the designated line of departure and the 
division's night positions before the divi- 
sion jumped off. The division launched its 
drive about 1330, an hour and a half late. 
The delay was largely because the 25th 
Marines, after being relieved by Army units 
in midmorning, had to move laterally about 
2,500 yards to take position on the right of 
the line. Once they jumped off, the ma- 
rines drove forward against very little re- 
sistance and by 1630 reached their objec- 

:l ° Ibid., App. G, pp. 2-3. 

41 NTLF Opn Order 22-44, 4 Jul 44. 

32 27th Inf Div G-3 Jnl, 4 Jul 44, Msg 4G. 

33 165th RCT Record of Opns Forager, p. 9; 
Love, Battle for Saipan, pp. 693-703. 



tive for the day, the 0-8a line, which was 
about 1,200 yards from the line of 
departure. 34 

The rapid and almost uncontested prog- 
ress was indicative of the total collapse of 
General Saito's plans for establishing a final 
defense line across the entire northern neck 
of Saipan. The 4th Marine Division had 
overrun the whole left flank of the pro- 
posed line. The 136th Infantry Regiment 
should have contested this ground, but 
whatever remained of that unit was scat- 
tered and isolated behind the American 
lines, mostly in the area around Radar 
Hill. 35 With the collapse of the enemy left, 
all that remained under Saito's control was 
the Navy sector and a thin slice of the 
135th Infantry's area. These were in the 
zone of the 27th Division. Even there, the 
defense was disorganized and confused. 
Japanese officers captured on the 5th re- 
vealed that their "front line units were 
mixed up, the communications were badly 
disorganized, . . . there was little or no or- 
ganized resistance at the present time, no 
organized supply plan and very little artil- 
lery, if any, remaining." 3e Yet to the 
Japanese military mind, disorganization, 
lack of supplies, and lack of communica- 
tions was no excuse for an abatement of 
effort. What the enemy lacked in the ordi- 
nary sinews of war he made up in deter- 
mination. As the 27th Division began to 
probe into Saito's last shattered defense 
line, the degree of that determination was 
made manifest. 

S4 4th Marine Div Opns Rpt Saipan, Annex I, 
24th RCT Rpt 3 p. 63; Ibid., Annex J, 25th RCT 
Rpt, p. 8. 

3B NTLF Rpt Marianas, Phase I, Incl D, G-a 
Rpt, pp. 53-55, 93- 

sr > Ibid., p. 54. 

2jth Division 

The newly designated line of departure 
for the 27 th Division ran cast from the 
beach just north of the village of Tanapag 
to a point just south of Hill 767. Facing 
this line from right to left were the 2d Bat- 
talion, 165th Infantry, 3d Battalion, 165th, 
3d Battalion, 105th, and 2d Battalion, 
105th. The terrain over which the division 
was to move was of two kinds. On the left 
in the zone of action of the 105th Infan- 
try, the ground was a low, slightly rolling, 
coastal plain. The most important land- 
mark on the plain was a large coconut 
grove about 600 yards east of Tanapag vil- 
lage. The main coastal road ran along the 
beach and was paralleled by a small cane 
railroad. Just above Tanapag, at Road 
Junction 2, the coastal road was joined by 
a cross-island highway. Just to the east of 
the coconut grove, the highway made, a 
U-shaped turn and from the north leg of 
the U, at Road Junction 64, another, 
smaller, road branched off, wound in a 
southeasterly direction through a canyon 
fifty to sixty feet deep, and came out into 
the hills below Hill 721, one of the two 
high points on the ridge that rose up from 
the plain in the center of the island. 
The canyon, winding uncertainly between 
steep, cave-studded cliffs on cither side, 
was soon to be called "Harakiri Gulch" by 
the men of the 27th. The floor of the 
gulch, never more than fifty yards wide, 
was covered with sparse undergrowth dot- 
ted with trees. In length, it ran about 400 
yards. 37 Lying athwart the main line of 
advance, the canyon was an ideal defensive 
position. From the west mouth of the 
gulch, running all way to the sea, was a 

Love, Battle for Saipan, pp. 715-16. 



Harakiri Gulch, sjth Division soldiers patrolling road following capture of 
this strongpoint on y July. 

deep dry gully that also provided ideal 
cover for enemy movement. 38 

K Company of the 165th Infantry drew 
first blood in the two-day fight for Hara- 
kiri Gulch. Soon after the jump-off, an ad- 
vance patrol climbed down the south face 
of the canyon but received such inhospit- 
able treatment from the Japanese below 
that the men climbed right back up again, 
dragging their wounded with them. Shortly 
thereafter two tanks started down into the 
gulch via the road to the west. Within a 

38 Historical Division, War Department, AMER- 
tions (Washington, 1946) (hereafter cited as 
AFAS, Small Unit Actions), p. 74, map on p. 99. 

few minutes both were disabled by Japa- 
nese who darted out from the ditches and 
placed mines on them. Three more tanks 
from the same platoon appeared over the 
edge of the precipice in an attempt to res- 
cue those below. After an hour and a half 
of maneuvering and firing, one of the 
stricken vehicles was recovered; the other 
had to be abandoned. 33 For the rest of the 
afternoon Company K made repeated stabs 
into the gulch, but each failed. Self- 
propelled mounts were sent down the road 
to search the caves on the north side with 

ao Appleman, Army Tanks in the Battle for Sai- 
pan, pp. 76-80. 



point-blank fire from their 75-mm. and 
105-mm. howitzers, but the infantrymen 
who followed found the going still too 
rough, and Captain Betts withdrew the 
company from the gulch and called for ar- 
tillery. Along the southern rim the entire 
3d Battalion, 165th, dug in for the night, 
tying in on the right with the 2d Battalion, 
165th, which had seen no significant action 
during the day. 40 

On the left of the 3d Battalion, 165th, 
Company L of the 1 05th was stopped in its 
tracks by fire from the opposite side of 
Harakiri Gulch and made no effort to force 
an entry into the canyon. To its left Com- 
pany K, 105th, tried to work its way into 
the coconut grove, but fire from the up- 
lands on the right interdicted the area, 
mortally wounding Captain Bouchard. The 
new company commander, 1st Lt. Roger 
P. Pcyre, then withdrew south of the 
grove. The 2d Battalion, 105th, had spent 
the day working its way slowly along the 
shore line and the coastal plain north of 
Tanapag. It had mopped up a series of 
small pillboxes, most of them abandoned, 
and had discovered a live mine field directly 
in the path of its advance. By the end of 
the day it had not quite reached its sched- 
uled line of departure, although the men 
had moved almost 800 yards through 
ground not previously rcconnoitered. 43 

6 July 
4th Marina Division 

Holland Smith's orders for 6 July called 
for the 27th Division to jump off at 0700 
in an effort to bring its line abreast of the 
marines on the right by ogoo. Assuming 

this would be accomplished on time, the 
4th Marine Division was to launch its at- 
tack at 0900, and the two divisions would 
continue to move abreast in a north- 
easterly direction toward the tip of the is- 
land, sweeping the remaining Japanese 
before them. 42 An hour or two after the 
27th Division had jumped off it became 
apparent to the corps commander that it 
was going to be impossible for it to keep 
pace with the marines. Consequently, at 
0900 General Smith changed his plans and 
assigned new missions. The 27th Division 
was to reorient the direction of its attack 
from northeast to north, thus assuming 
responsibility for about 2,600 yards of 
coastal strip from just above Tanapag to 
just above Makunsha, as well as for the 
first high ground immediately inland from 
the beaches, Harakiri Gulch, and Paradise 
Valley. The entire remainder of the island 
northeast of this sector was to be taken 
over by the 4th Marine Division. Once the 
right flank of the Army division reached its 
objective on the west coast just above 
Makunsha, it would be pinched out. 43 

To take over his newly expanded front, 
General Schmidt put all three of his Ma- 
rine regiments into the line — 25th, 24th, 
and 23d Marines from right to left. Ac- 
companied by thirteen tanks, the 25th 
Marines made fairly rapid progress north 
along the cast coast of the island, mopping 
up isolated Japanese troops and civilians 
in the many caves and cliffs that bordered 
the ocean. In this work the 25th was as- 
sisted by naval vessels, whose flat trajec- 
tory fire was ideally suited to the coastal 
targets. By midafternoon Mount Petosu- 

Love, Battle for Saipan, pp. 717-23. 
Ibid., pp. 728-^5. 

42 NTLF Opn Order 23-44, 5 J ut 44- 
4:1 NTLF Rpt Marianas, Phase I, Incl E, C-3 
Rpt, G-3 Periodic Rpt 22. 



kara was occupied, and the two assault 
battalions dug in for the night on either 
side of that elevation. Just before dark a 
group of from seven to eight hundred 
civilians came into the lines of the 25th and 
surrendered. Meanwhile, to the rear, the 
reserve battalion in mopping up a by- 
passed hill flushed a sizable covey of 
Japanese soldiers and killed sixty-one in a 
brief but lively fire fight. 44 

In the division center, the 24th Marines 
registered a day's gain of 1,400 to 1,800 
yards against sporadic resistance, 45 

On the left, the 23d Marines encoun- 
tered considerably more difficulty. Having 
been in reserve in the morning when it re- 
ceived its orders, the unit had to march 
some 4,300 yards before reaching its line 
of departure. Jumping off at 141 5, it soon 
came upon the cliff line that rimmed Para- 
dise Valley on the east. Here, the regiment 
came under enemy fire from caves well 
concealed by dense underbrush. As the ma- 
rines pushed down the slopes into the val- 
ley, hidden enemy machine guns and knee 
mortars opened up from the rear. With 
only an hour of daylight remaining, the 
regimental commander decided it was im- 
possible to continue the attack, and at 1730 
pulled his men back to establish defensive 
positions for the night on the high ground. 
There, the 23d Marines tied in with por- 
tions of the 27th Division but was entirely 
out of contact with the 24th Marines on 
the right. 46 

44 4th Marine Div Opns Rpt Saipan, Annex J, 
25th RCT Rpt, p. 9; Hoffman, Saipan, pp, 

5230-C! I, 

15 4th Marine Div Opns Rpt Saipan, Annex I, 
34th RCT Rpt, p. 24; Hoffman, Saipan, p, 519. 

4(1 4th Marine Div Opns Rpt Saipan, Annex H, 
23d RCT Rpt, p. 47; Hoffman, Saipan, pp. 

The Battle for Tanapag Plain 

On the morning of 6 July the 27th In- 
fantry Division was still on the near side of 
Harakiri Gulch and still short of its line of 
departure on the plain north of Tanapag 
village. On the line from right to left were 
Major Claire's 2d Battalion, 165th In- 
fantry, Major Mahoney's 1st Battalion, 
165th, 47 Colonel Bradt's 3d Battalion, 
105th, and Maj. Edward McCarthy's 2d 
Battalion, 105th. The plan for the day, as 
revised by General Smith's order of 0900, 
called for the 2d Battalion, 165th, to push 
toward the coast above Makunsha by way 
of Paradise Valley, On its left the 1st Bat- 
talion, 165th, and the 3d Battalion, 105th, 
were to rout the enemy still entrenched in 
Harakiri Gulch and then proceed north- 
ward. Finally, to the 2d Battalion, 105th, 
was given the job of pushing up the 
coastal plain to a point just south of 
Makunsha. 48 

On the right the division made no prog- 
ress in the effort to push through Paradise 
Valley. Capt. William J. Smith, command- 
ing Company F, 165th Infantry, tried to 
force his way into the valley by the trail 
that ran along its floor, but the hail of fire 
that greeted this effort discouraged him 
and he withdrew his men. After a futile 
effort to rout the enemy with tanks and 

47 This unit had taken over the positions oc- 
cupied on the 5th by the 3d Battalion, 165th 

48 The account of action of the 27th Infantry 
Division on 6 July, unless otherwise noted, is de- 
rived from Love, Battle for Saipan, pages 738— 
838, and AFAS, Small Unit Actions, pages 69- 
118. These accounts were written largely from 
interviews conducted shortly after the action by 
Captain Love, the historian assigned to the divi- 
sion. The official records are sparse to the point of 
being almost useless. This is particularly true of 
the records of the 105th Infantry, most of which 
were destroyed by fire. 



MAP 12 

self-propelled mounts, the whole battalion 
fell back to the western base of Hill 767 
and dug in for the night. 

The attack on Harakiri Gulch met with 
no more success. Jumping off about noon 
the 1st Battalion, 165th Infantry, at- 
tempted, as had the 3d Battalion the day 
before, to assault the canyon frontally, 
moving perpendicularly to its axis. In the 
course of this effort the men of Company 
A witnessed an incident that was to give 
the name to the area. Following an intense 
ten-minute mortar preparation, the com- 
pany proceeded slowly into the valley and 
was greeted by a scries of explosions that 
forced the lead platoon to duck for cover. 
When the fireworks had abated about fif- 

teen minutes later, the men investigated a 
group of straw shacks located on the sides 
of the gulch in the path of their advance. 
In each of these they found groups of three 
or four Japanese soldiers who had com- 
mitted suicide by pressing hand grenades 
to their abdomens. All together, about 
sixty of the enemy were discovered to have 
ended their lives in this fashion. Neverthe- 
less, fire from the gulch below continued 
intermittently throughout the afternoon, 
and by evening Major Mahoney's battalion 
abandoned all thought of further advance 
and dug in again on the rim overlooking 
the gorge. 

On the western flank of the gulch, the 
3d Battalion, 105th Infantry, was equally 



{Map is) 

In this area, most 
bat fell on Capt. 

of the burden 01 com 
Robert J. Spaulding's Company L. During 
the morning Spaulding made two separate 
attempts to get across the gulch. He or- 
dered his i st Platoon, on the right, to crawl 
down into a small tributary draw that 
branched off from the main gulch in a 
southwesterly direction. The platoon was 
to work down the draw to its mouth and 
there set up machine guns that could cover 
enemy positions on the floor of the gulch 
and fire into the caves on the opposite side. 
Under cover of this support, Spaulding 
proposed to send his 2d Platoon over the 
near walls of the canyon, across the floor 
of the gulch, and up the opposite side. He 
also had at his disposal a skeleton platoon 
of four light tanks that he intended to send 
up the gulch along the trail that entered 
it from the east. 

Company L moved off to the attack at 
0700. The 1 st Platoon crawled up over the 
ridge and down into the tributary ravine 
without drawing fire. Moving stealthily in 
single file along this narrow corridor, the 
platoon escaped detection and reached the 
corridor's mouth. There, the men set up 
two light machine guns and began firing 
at the caves in the face of the opposite wall, 
only to be greeted by return fire from the 
disabled American tank that had been left 
in the gulch the day before and was now 
in the hands of the Japanese. Meanwhile, 
the four light tanks had arrived, and 
Spaulding ordered them to work up the 
trail that ran through the middle of the 
gulch. An infantryman, Pfc. James R. 
Boylcs, volunteered to accompany the but- 
toned up lead tank to guide it, but he was 
soon killed and thereafter no direct com- 
munication could be maintained between 
tanks and infantrymen. To add to the con- 

fusion, three enemy soldiers then jumped 
out of the bushes and clapped a magnetic 
mine onto the side of the third tank in line, 
disabling it. Eventually the crew from 
the crippled tank was evacuated, and the 
tank platoon commander, 2d Lt. Gino 
Ganio, was able at last to get well up into 
the gulch and spray the walls on the north 
side. Nevertheless, by this time (noon) 
Company L had withdrawn again to the 
rim of the gorge, and no further effort to 
breach the canyon was made on the 6th. 

Meanwhile, Company K, 105th Infan- 
try, commanded by Lieutenant Peyre, was 
having its own troubles in the area of the 
coconut grove in the valley below and to 
the west of the mouth of Harakiri Gulch. 
Jumping off on schedule at 0700, Peyre's 
men moved along a deep gully that circled 
the south edge of the grove, making use 
of the cover and concealment it offered. 
Once they emerged from the ditch, how- 
ever, they were taken under fire by Japa- 
nese machine guns located near the center 
of the grove, and the whole company was 
pinned down. At this juncture a platoon of 
five light tanks commanded by 1st Lt. 
Willis K. Dorey hove into view and within 
ten minutes cleared the way for the in- 
fantrymen to move into the grove. For 
about an hour the men of Company K 
worked their way among the stock piles of 
enemy supplies that abounded in the area 
and by 08 1 5 reached the north edge of the 
grove, facing the open ground beyond the 
cross-island road. 

Once his troops arrived at this objective, 
Lieutenant Peyre ordered Dorcy's tanks to 
move along the road until they reached po- 
sitions from which they could put effective 
fire on the cliffs to the right front. So long 
as the tanks were firing the infantrymen 
were able to move about the grove at will, 



but whenever the tankers ceased fire the 
Japanese in the cliffs opened up again. Un- 
fortunately too, at this point, tank-infantry 
communications failed, and Peyre could 
neither reach Dorey by radio nor attract 
his attention with hand signals. Conse- 
quently the tank commander merely kept 
patrolling the road, laying down a blanket 
of fire on the cliffs, until about 1000 when 
the tanks ran out of ammunition and had 
to retire. 

Pcyre dug in as well as he was able to 
await the tankers' return. A hundred yards 
ahead of him in the open terrain north of 
the grove was a small knoll on which were 
located three enemy machine gun positions. 
To interdict these as well as the remaining 
guns on the cliffs to his right, Company 
K's commander brought his own machine 
guns to the north edge of the grove and 
prepared to launch an attack against the 

When the tanks returned at 1030, Lieu- 
tenants Peyre and Dorey conferred and 
laid their plans. The right platoon of 
Company K would move ahead to capture 
the knoll under cover of fire from the left 
platoon. Dorey, with his tanks, would 
again proceed up the cross-island road, take 
the trail that led into Harakiri Culch, and 
neutralize the enemy fire in the cliffs. 

The lead platoon jumped off about 1 045 
and was immediately met by a deadly hail 
of small arms and machine gun fire that 
forced the men to take to the earth. 
Lieutenant Peyre, seeing his right platoon 
stalled, ordered his left platoon to try for 
the rise. Just as these men were venturing 
out of the coconut grove, the Japanese 
counterattacked down the cliffs and along 
the paths that led to a gully just behind 
the rise of ground that was the American 
objective. Total chaos ensued as a result 

of a tremendous explosion that sent bodies 
and limbs of the leading Japanese into the 
air in all directions. Apparently, one of the 
enemy had stepped on the horn of an em- 
bedded sea mine, thus setting off a series 
of mines scattered over the area. Whatever 
the cause of the explosion, it created havoc 
among the Japanese and abruptly stopped 
the counterattack. In the American lines 
the results were not so serious, and al- 
though a few men were wounded by flying 
debris, the effect of the concussion was 

Meanwhile, orders had come down for 
Company G, 105th Infantry, to relieve 
Company K in the coconut grove. After 
receiving General Smith's orders indicating 
that the 27th Division would change the 
direction of its attack from northeast to 
north toward the coast line, General Griner 
had decided to shift the emphasis of his 
division attack from the left to the right of 
his line. Hence, to bolster the efforts 
of the 3d Battalion, 105th, against Harakiri 
Gulch, he ordered Company G to relieve 
Company K so that the latter could move 
out of the coastal plain and into the reserve 
area behind Us parent unit. For the rest of 
the afternoon the area along the coast 
would be assigned entirely to the 1st and 
2d Battalions, 105th Infantry. 

It was the second of these battalions, 
commanded by Major McCarthy, that had 
been responsible for the area immediately 
abutting the scacoast during the morning. 
As day broke McCarthy had Companies E 
and F drawn up in a tight perimeter 
around Road Junction 2. Directly ahead 
athwart his line of advance, was the mine 
field, discovered the day before, that ran 
from the coastal road to the railroad and 
that was about 250 yards in depth. It con- 
sisted of about 150 Japanese general pur- 



Waiting To Move Up north of the coconut grove on 3 July. 

pose bombs set in the ground in four rows, 
noses up. Only about a hundred had been 
fuzed. 41 * Immediately beyond the mine 
field was the gully that ran down to the 
sea from the western mouth of Harakiri 
Gulch. To the right (east) of McCarthy's 
bivouac area was a wide expanse of open, 
slightly rolling ground, which was covered 
by small arms and automatic weapons fire 
from the cliffs still farther to the east. 

The 2d Battalion's commander decided 
to move his men along the narrow strip 
of beach between the road and the lagoon 
in order to avoid the mine field. To elimi- 
nate the series of pillboxes strung along the 

49 27th Inf Div G-a Periodic Rpt 20, 6 Jul 44. 

shore in this area, he called for a rolling 
artillery barrage in advance of the infantry. 
Company F was to take the lead, to be 
followed by E Company, which would fan 
out to the right once the far edge of the 
mine field was passed. 

Promptly at 0700 Company F jumped 
off and within a few minutes had reached 
the northern limit of the mine field. At this 
point it received a heavy burst of machine 
gun and small arms fire from its direct 
front. McCarthy at once tried to put in a 
call for tanks and self-propelled mounts, 
but discovered that his radio communica- 
tions were out; he then sent a runner to 
order up the vehicles. This involved a trip 
all the way back to Tanapag. 



Meanwhile, the men of Company E had 
managed to crawl to the right along the 
north edge of the mine field and to deploy 
in a three-platoon front along a line run- 
ning east of the coastal road. About 0900 
Major McCarthy decided to withdraw 
Company F from its cramped positions be- 
tween the road and the beach and send it 
around to the right of Company E to close 
the gap between the 2d and 3d Battalions. 
This move took about an hour. Also, Com- 
pany A of the i02d Engineer Combat 
Battalion was brought forward to deactiv- 
ate the mine field. 

At 1000 five medium tanks commanded 
by 1 st Lt. Dudley A. Williams of the 76ad 
Tank Battalion put in their appearance at 
Road Junction 2. Rather than send them 
down the road, which he believed was al- 
most certainly mined, McCarthy ordered 
them to proceed single file along the rail- 
road track to the right. The lead tank un- 
fortunately snarled its tread in the steel 
rails and became immobilized. While an 
effort was being made to clear a path 
through the mine field so that the second 
tank could be worked around the first, the 
enemy opened fire, scoring direct hits on 
both tanks. Lieutenant Williams hooked 
cables to the two vehicles and hauled them 
loose of the tracks and clear of the area 
before any more damage was done. 

By this time it was apparent that the 
chief source of enemy fire came from the 
gully in front of the mine field, and Com- 
any E sent out a squad to rush the gully 
and knock out the machine gun position 
that seemed to be causing most of the 
trouble. The squad leader got as far as the 
gully and located the position in question, 
but was wounded and had to withdraw 
before he could eliminate it. 

By midafternoon the entire 2d Battalion, 
105th Infantry, appeared to have bogged 
down. Companies E and F were facing the 
gully just north of the mine field and Com- 
pany C was still at the north edge of the 
coconut grove. Anxious to get on with the 
day's business, General Griner at 1520 or- 
dered the regimental commander, Colonel 
Bishop, to commit his reserve, the 1st 
Battalion, 105th, commanded by Colonel 
O'Brien. Bishop objected to committing his 
reserves at such a late hour and argued 
that an attack would not give sufficient 
time before dark for the front-line troops to 
prepare a proper perimeter defense. so His 
objection was overruled. On Griner's or- 
ders, O'Brien's unit was to be inserted on 
the right flank of the 2d Battalion, 105th, 
and from that point was to drive north to 
Makunsha village on the shore before 
nightfall. 51 

Even before this move could be exe- 
cuted, relief to the men immediately in 
front of the enemy-infested gully came 
from a different quarter. About 1530 
Lieutenant Dorey, after refueling and re- 
supplying, arrived on the scene with two 
other tanks in addition to his own. Observ- 
ing that the infantry was apparently com- 
pletely immobilized, he drove straight into 
the troublesome gully pushing the enemy 
before him and slaughtering them with 
canister and machine guns. For half an 
hour he kept this up, killing about 150 
Japanese in the gully and literally paving 
with dead bodies the way for a renewed 
advance of the 2d Battalion, In the course 
of this action, Japanese soldiers, armed 
with magnetic mines, attacked one of the 

50 Ltr, Bishop to Gen A. C. Smith, 2^ Feb 5^, 

51 27th Inf Div C-3 Jnl, 6 Jul 44, Msg ;}G. 



light tanks and it lost its track. Tn spite 
of his valiant efforts, Dorey was unable to 
rescue cither the damaged tank or its 
crew/' 2 

Meanwhile, back at the regimental com- 
mand post, Colonel Bishop was outlining 
his plans for the final move up the coast to 
Makunsha. As directed by the division 
commander, O'Brien's ist Battalion was to 
move into line between the other two bat- 
talions of the 1 05th. To make room for this 
maneuver, Company F was to move back 
around the rear of Company E to the left 
of the regimental line where it would again 
take up a position between the railroad 
track and the beach. Company G, com- 
manded by Captain Olander and still in 
the coconut grove, was to be attached to 
the 3d Battalion and swing on its right 
flank across the western mouth of Harakiri 
Gulch in order to bottle up the enemy 
there. Such a movement would presum- 
ably protect the rear of the ist Battalion, 
and the next morning the rest of the 3d 
Battalion could mop up the enemy iso- 
lated in the gulch. 

Pursuant to these instructions, Colonel 
O'Brien brought his battalion into line, 
with Company B on the right, A on the 
left, and C echeloned to the right rear. His 
apprehension over the role assigned to his 
men was apparent to Captain Ackerman, 
A Company commander, who later testi- 
fied: "Obie was nervous and restless, as 
usual. He drew a picture for us and told 
Dick [Capt. Richard F. Ryan, of Com- 
pany B] and I that no matter what else 
happened, we were to keep going. 'Its the 
old end run all over again. Whenever they 
got a job nobody else can do, we have to do 

5a Appleman, Army Tanks in the Battle for Sai- 
pan, pp. 91-92. 

it. Sooner or later we're going to get caught 
and this may be it.' " S3 

The ist Battalion was in line by approx- 
imately 1645, following F Company's shift 
to the division left flank along the beach. 
Between that time and 17 15 both battal- 
ions resupplicd and organized their lines. 
At 1715 the 105th In fantry move d off in 
a co-ordinated attack. 54 |(A/a/> 79)] 

On the left of the line, the 2d Battalion 
had little difficulty moving ahead in the 
wake of the devastation caused by Lieuten- 
ant Dorey's tanks. Although Company F 
delayed slightly to investigate a series of 
Japanese pillboxes along the beach, by 
1800 the whole battalion had advanced 
about 600 yards. At that point it built up 
its perimeter for the night. O'Brien's bat- 
talion ran into more trouble. On reaching 
the gully, Company A encountered a nest 
of fifteen to twenty Japanese. Some were 
wounded and some were still trying to hide 
from Dorey's tank fire by hugging the 
walls of the trench on the near side. Acker- 
man's men waded in with bayonets and 
knives and after a 30-minute hand-to-hand 
fight, cleaned out the pocket. Once across 
the gully, Company A rushed headlong 
some 500 yards in spite of increasingly 
heavy machine gun fire from the cliffs to 
the right. This fire was falling even more 
heavily on Company B and succeeded, 
among other things, in killing the company 
commander, Captain Ryan, who was re- 
placed by ist Lt. Hugh P. King. Mean- 
while, on the battalion right rear, Com- 
pany C was faced with the same machine 
guns cmplaccd on the knoll north of the 
coconut grove that had previously stopped 
both K Company, 165th, and G Company, 
105th. For the rest of the day and even 

53 Quoted in Love, Battle for Saipan, p. 810. 

54 105th RCT Jnl, 6 Jul 44, p. 7a. 



MAP 13 

after dark, Company C battled to take 
out these positions. Not until a self- 
propelled mount was finally brought in to 
wipe them out was the entire company able 
to rejoin the rest of the battalion in its 
night perimeter. 

Meanwhile, immediately to the right 
Captain Olander of Company G, 105th, 
was trying to carry out his mission of seal- 
ing up the western mouth of Harakiri 

Gulch. Working its way along the road that 
led into the gulch the lead platoon, just 
before dark, stumbled onto a nest of Jap- 
anese. A brisk hand-to-hand fight ensued, 
with inconclusive results, and, in view of 
the lateness of the hour, the company com- 
mander ordered all of his men to pull back 
west along the road to the point from 
which they had started. There Captain 
Olander called Colonel Bradt, to whose 



battalion he was now attached, and ad- 
mitted his inability to build up a line across 
the mouth of the gulch. He was given per- 
mission to dig in on the high ground over- 
looking Road Junction 64, from which 
point he hoped to be able to interdict 
movement from the gulch with machine 
gun fire. 

By nightfall then, the 1st and 2d Bat- 
talions, 105th (less Company G), were 
digging in in positions about 900 yards 
northeast of Road Junction 2. On the left, 
Companies E and F bivouacked in a tight 
perimeter between the road and the rail- 
road. The narrow corridor between the 
road and the beach was outposted by two 
men armed with carbines and with express 
orders to report any signs of enemy move- 
ment along the beach. Inside the perimeter 
were stationed the mortars of both rifle 
companies as well as the mortars and ma- 
chine guns of the heavy weapons company. 
In placing his heavy weapons and machine 
guns, Major McCarthy assumed that be- 
fore nightfall the 1st Battalion would have 
reached its objective north of his perime- 
ter. Hence, without neglecting the north- 
ern approaches to his position altogether, 
he concentrated his defenses on the eastern 

Meanwhile, Colonel O'Brien's 1st Bat- 
talion had come abreast. Rather than push 
on to the beach ahead of the ad Battalion 
as originally planned, O'Brien, after con- 
sultation with McCarthy, decided to tic in 
for the night to the right of the 2d Bat- 
talion. His perimeter was drawn up in the 
shape of an arc whose terminal points 
rested on the railroad just east of the right 
side of the 2d Battalion's perimeter. This 
arrangement meant that two whole pla- 
toons of the 2d Battalion were now in- 

side the final perimeter. More important, 
O'Brien's perimeter screened one of the 2d 
Battalion's antitank guns as well as all of 
Company H's heavy machine guns, which 
had been emplaced so as to protect the 
eastern leg of McCarthy's original perim- 
eter. Thus, by hedging in the 2d Battalion 
from the east, O'Brien in effect subtracted 
from the combined fire power of the two- 
battalion perimeter. 

Even more significant was the fact that 
between the 1st Battalion, 105th, and 
Company G, 105th (attached to the 3d 
Battalion), lay a gap of about 500 
yards. However, the ground was open and 
O'Brien took the precaution of placing all 
of his antitank guns in such a position as 
to bear directly on the gap. By the time all 
these arrangements were completed it was 
well after dark. The morning, it was hoped, 
would bring the 105th Infantry to its ob- 
jective line at the shore and an end to its 

7 July 

Banzai Attack 

About an hour after dark, an American 
soldier patrolling the road in the vicinity 
of the command post of the 3d Battalion, 
105th Infantry, came upon a lone, armed, 
Japanese lying asleep. He forthwith took 
him prisoner and sent him back to head- 
quarters for interrogation. The Japanese 
proved to be a "leading seaman" of the 
Jtjth Guard Force, and his testimony, re- 
luctantly given, was sufficient cause for 
deep alarm. An all-out attack by the entire 
remaining Japanese force on the island, he 
said, had been ordered for the night of 6-7 
July. Word was immediately sent out to 
all major units of the division as well as to 

the capture of northern saipan 


Holland Smith's headquarters to prepare 
for the worst.'™ 

In the front line below Makunsha, 
Colonel O'Brien and Major McCarthy 
went into conference on receiving this in- 
formation. Both were worried about the 
gap that extended some 500 yards south- 
eastward to the night positions of Company 
G, 105th Infantry. O'Brien called the regi- 
mental command past and asked for re- 
inforcements to fill the gap but was told 
that none were available. Colonel Jensen, 
the regimental executive, in turn called for 
help from division headquarters. He too 
received a negative answer. The two bat- 
talion commanders would have to make 
out with what they had on hand.' 1 ' 1 

The Japanese counterattack that was 
now mounting had in fact been ordered 
early on the morning of the 6th. At 0600 
General Saito had issued his final proc- 


I am addressing the. officers and men of the 
Imperial Army on Saipan. 

For more than twenty days since the Amer- 
ican Devils attacked, the officers, men, and 
civilian employees of the Imperial Army and 
Navy on this island have fought well and 
bravely. Everywhere they have demonstrated 
the honor and glory of the Imperial Forces. 
I expected that every man would do his duty, 

Heaven has not given us an opportunity. 
We have not been able to utilize fully the 
terrain. We have fought in unison up to the 
present time but now we have no materials 
with which to fight and our artillery for at- 
tack has been completely destroyed. Our 
comrades have fallen one after another. Des- 
pite the bitterness of defeat, we pledge, 
"Seven lives to repay our country." 

The barbarous attack of the enemy is being 
continued. Even though the enemy has oc- 
cupied only a corner of Saipan, we are dying 
without avail under the violent shelling and 
bombing. Whether we attack or whether we 
stay where we are, there is only death. How- 
ever, in death there is life. We must utilize 
this opportunity to exalt true Japanese man- 
hood. I will advance with those who remain 
Lo deliver still another blow to the American 
Devils, and leave my bones on Saipan as a 
bulwark of the Pacific. 

As it says in the "senjinkum" [Battle 
Ethics], "I will never suffer the disgrace of 
being taken alive," and "I will offer up the 
courage of my soul and calmly rejoice in liv- 
ing by the eternal principle." 

Here I pray with you for the eternal life 
of the Emperor and the welfare of the coun- 
try and I advance to seek out the enemy. 
Follow rne! r,T 

Actually, General Saito was too feeble 
and sick to lead the charge in person. 
Shortly after issuing his final order he com- 
mitted suicide. A captured Japanese officer 
who was with the general almost until the 
end described what probably took place: 
"Cleaning off a spot on the rock himself, 
Saito sat down. Facing the misty east 
saying 'tenno heika! banzai! [Eong 
live the Emperor] ... he drew his own 
blood first with his own sword and then 
his adjutant shot him in the head with a 
pistol." 58 

The exact number of Japanese to partic- 
ipate in the attack is unknown. A count 
taken later on the authorization of General 
Griner revealed 4,311 enemy dead in the 
area covered by the attackers, although un- 
doubtedly some of these had been killed 
by naval gunfire or artillery before the ban- 
zai charge got under way/'" A captured in- 

55 27th Inf Div G-3 Jnl, 6 Jul 44, Ms$? 57; 
27th Inf Div G-2 Periodic Rpt 21, 7 Jul 44. 
,r,,! Love, Battle for Saipan, pp. 858-59. 

5t NTLF Rpt Marianas, Phase I, Incl D, G-a 
R Pt, PP- 57-58. 

58 Ibid., App. G, p. 3. 

59 Buckner Bd Rpt, Exhibit FFF, Ltr, CG 27th 
Inf Div to CG NTLF, r6 Jul 44. 



tclligence officer of the 43d Division at first 
estimated that the total Japanese force 
came to no more than 1,500, but later re- 
vised this upward to 3,ooo. co Another 
prisoner of war, a Korean laborer, also gave 
3,000 as the approximate number," 1 and 
this is probably as acceptable an estimate 
as any. 8a 

The truth is that even the Japanese com- 
manders themselves had no very clear pic- 
ture of the number of men left under them. 
The attacking force was drawn from 
almost every conceivable unit on the island, 
forming a composite group of stragglers. 
Specific identifications made among Jap- 
anese dead included the 118th, 135th, and 
136th Infantry Regiments, 43d Division 
headquarters and 43d Field Hospital, 3d 
Independent Mountain Artillery Regiment, 
16th Shipping Engineer Regiment, and 
sundry naval units including combat, 
maintenance, and labor personnel. Many 
of the Japanese were poorly armed with 
rusty rifles and some merely carried poles 

«" Ibid., Annex II, p. 7 ; Hq FEC, Mil Hist Sec 
Special Staff, to Chief of Mil Hist, 5 Jun 5a, Incl, 
Comments by Maj. Takashi Hiragushi. Major Hi- 
ragushi, was taken prisoner after the counterattack 
of 7 July. For reasons unknown he assumed the 
name of Maj. Kiyoshi Yoshida, the 31st Army 
intelligence officer, who had been killed in action. 
The NTLF G-2 Report refers to this officer as 
Yoshida. His true identity was not revealed until 
after the war. 

el NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, p. 96. 

1,2 General Holland Smith's final estimate lay 
between 1,500 and 3,000 (Smith, Coral and Brass, 
p. 195). General Griner believed the number was 
not less than 3,000 and probably more (CO 27th 
Irif Div to CG NTLF, 16 Jul 44, Iluckncr Board 
Rpt, Exhibit FFF). A special hoard, appointed 
by Admiral Spruancc to survey the circumstances 
surrounding the counterattack, estimated the 
number of enemy involved to lie between 2,500 
and 3,000 (Comdr Fifth Fleet, Rpt of Japanese 
Counterattack at. Saipan on 7 Jul 44, 19 Jul 44). 

to which crude knives and bayonets were 
attached. 63 

Poorly armed or not, the impact of this 
horde was overwhelming. In the words of 
Major McCarthy, one of the few officers 
to survive it, "It was like the movie stam- 
pede staged in the old wild west movies. 
We were the cameraman. These Japs just 
kept coming and coming and didn't stop. 
It didn't make any difference if you shot 
one, five more would take his place. We 
would be in the foxholes looking up, as I 
said, just like those cameramen used to be. 
The Japs ran right over us." ° 4 

About 0400 on 7 July the main body 
of the desperate attackers started south 
from Makunsha between the shore line and 
the base of the cliffs bordering Tanapag 
plain. Although there is no evidence that 
the movement was organized, the mount- 
ing flood sluiced out along three principal 
channels. The main group charged down 
the railroad track, hitting the American 
perimeter below Makunsha; another at- 
tacked positions of the 3d Battalion, 105th, 
at Harakiri Gulch; the attackers facing the 
gap between these two American positions 
continued through unopposed. 05 

Shortly before 0500 the full force of the 
attack struck the perimeter of the 1st 
and 2d Battalions, 105th Infantry, and in 
twenty-five minutes of fierce close-quarter 
fighting the American positions were over- 
run. During the first moments of impact 
Colonel O'Brien again made himself con- 
spicuous by his fortitude. With a pistol in 
each hand he joined battle with the 

,i:! NTLF Rpt Marianas, Phase I, Incl D, G-2 
Rpt, pp. 59-60. 

61 Quoted in Love, Battle, for Saipan, pp. 

60 Unless otherwise noted, the account of the 
banzai charge is derived from Love, Battle for 
Saipan, pages 871-92. 



deluge, firing until his magazines were 
empty. Then, though seriously wounded, 
he manned a .50-caliber machine gun and 
kept firing until killed. 66 With him went a 
good percentage of the officers and men of 
both battalions. 

The tide rolled on, and before it stum- 
bled most of the survivors of the perimeter. 
Among those left behind was Sgt. Thomas 
A. Baker of Company A. Although severely 
wounded, he refused to let himself be car- 
ried back with the retreat. Preferring cer- 
tain death to further risking the lives of his 
comrades he demanded to be left, armed 
only with a loaded pistol. When his body 
was later discovered the gun was empty 
and around him lay eight dead Japanese. 67 

Meanwhile, the left flank of the enemy 
had swiftly penetrated the gap. One group 
of Japanese spread out to attack the 3d 
Battalion, 105th, but from their dominat- 
ing positions on the high ground above 
Harakiri Gulch the men of the 3d Bat- 
talion were able to repulse the attack and 
hold their positions intact. 68 Another group 
hit the 3d Battalion, 10th Marines (Ar- 
tillery), which had set up its guns the day 
before about 500 yards southwest of Tan- 
apag village. Only one of the batteries (H) 
was in a position to fire and it succeeded, 
among other things, in knocking out a Jap- 
anese tank before the men were forced to 
retreat pell-mell, leaving the breechblocks 
and firing locks in their howitzers. The ma- 
rines of Battery T, after expending all of 
their small arms ammunition, removed the 

'''' For this and other notable demonstrations 
of bravery on Saipan, O'Brien was posthumously 
awarded the Medal of Honor (WDGO 35, 9 May 

' W45 ) ■ 

U7 For this action, Baker was posthumously 
awarded the Medal of Honor (WDGO 35, ij May 

1 945 ) ■ 

es 105th RCT Opns Rpt Forager, p. 10, 

firing locks from their howitzers and fell 
back south along the railroad track to the 
positions of Battery G, where the two units 
held fast until relieved that afternoon by 
the 106th Infantry. 69 

Meanwhile, the 27th Division artillery 
was pouring as many shells into the enemy 
as could safely be done without endanger- 
ing the retreating American troops. Be- 
tween 0515 and 0615 the three light bat- 
talions expended a total of 2,666 rounds in 
the zone of action of the 105th Infantry. 
This represented an average of more than 
forty rounds a minute. 70 

By the time the men of the two advanced 
battalions had retreated as far as the 
northern edge of Tanapag village, they ran 
into the van of the left prong of the Jap- 
anese force, which had come through the 
gap, past the positions of the 10th Marines, 
and then gone on to the command post of 
the 105th Infantry, where the attack was 
finally stopped. At this point, two officers, 
Captain White of Company F and Lieu- 
tenant King of Company B, rallied the re- 
treating men and brought some organiza- 
tion out of the confusion. They were able 
to persuade most of the troops to take 
cover in Tanapag village. While directing 
this diversion, King was killed. Meanwhile 
Major McCarthy, the 2d Battalion com- 
mander, had come up, and with the help 
of other surviving officers and noncom- 
missioned officers he was able to organize 
a perimeter within Tanapag village by 
about 0800, three hours after the initial 
attack had been made. 71 

011 Hoffman, Saipan, pp. 234-25. 

10 27th Inf Div G-3 Periodic Rpt 21, 7 Jul 44, 
p. 2. 

71 This account is derived from Love, Battle for 
Saipan, pages 876-932. 



Marines Examining Destroyed Enemy Tank,, which was knocked out by a 
battery from the iolh Marines during the Japanese banzai attack. 

For the next four hours the beleaguered 
men fought a bitter house-to-house battle 
with the Japanese that had surrounded 
and were infiltrating the village. The 
Americans were out of communication 
with the command posts to the rear, short 
of ammunition and water, and had no 
means of evacuating or properly caring for 
their wounded. Shortly after noo Mc- 
Carthy tried to lead a small force back to 
the regimental command post to bring up 
help for the wounded. Just as he got un- 
der way, his group was hit by two concen- 
trations of American artillery and those 
men who were able to do so stampeded into 
the water and swam for the reefs. Some of 
these returned to establish another small 
perimeter below Tanapag, where they re- 
mained out of touch with the main body of 
their regiment's troops in the village itself 

and the command post, which was still far- 
ther to the rear. 

Finally, shortly after noon, the first sign 
of relief appeared in the form of a platoon 
of medium tanks that rolled down the road 
from the direction of the command post. 
The vehicles fired indiscriminately at areas 
that might be presumed to contain enemy 
troops, but because there were no commun- 
ications between the tanks and the infantry 
there could be no co-ordinated effort to 
route the enemy or rescue the surrounded 
troops. Finally, McCarthy was able to get 
to the lead tank, and climbing in himself, 
lead a group of about thirty-five of his men 
back down the road, reaching the regi- 
mental command post by about 1500. 
Under his persuasion a convoy of trucks 
and DUKW's, loaded with medical sup- 
plies and ammunition, was dispatched 



toward Tanapag village. Some of the ve- 
hicles were knocked out en route, hut three 
got through and returned fully loaded with 
wounded. Still later, a group of LVTs of 
the 773d Amphibian Tractor Battalion was 
sent by water to rescue some of the men 
still fighting on the beach or stranded on 
the reefs. Others had already swum out 
and had been picked up by naval landing 
craft to be carried to destroyers waiting 
outside the reef. About 2200 the last sur- 
vivor left the perimeters in Tanapag village 
and below. All together, out of the 1st and 
2d Battalions, 105th Infantry (less Com- 
pany C), 406 officers and men were 
killed and 5 1 2 wounded. 72 

American Count ermeasures 

Back at 27th Division and corps head- 
quarters, word of the Japanese banzai 
charge was gradually filtering through. In 
response to the news, General Grincr at 
0920 ordered Colonel Stebbins to com- 
mit his 1 06th Regiment into the line 
and attack northeast astride the railroad 
track. 73 About the same time corps at- 
tached the 3d Battalion, 6th Marines, to 
the 27th Division, and at 1050 the battal- 
ion was ordered to mop up an enemy force, 
reported to be 100 strong, in the Tanapag 
area. 74 At 11 00 Grincr requested that 
some Marine tanks be released to the di- 
vision from corps control, but this was re- 
fused. 7 " According to the 27th Division 
commander, "headquarters did not accept 
my version of the importance of the action 

then in progress." 7,J However, not long 
afterward, Holland Smith did order the 
two Marine divisions to release 1,000 
rounds of 105-mm. howitzer ammunition 
to the Army division, which by now was 
running short. 77 

By 1000 Colonel Stebbins had the 106th 
Infantry in line with the 1st Battalion on 
the left and the 2d Battalion on the right 
of the railroad track. 78 They moved for- 
ward slowly. The 1st Battalion met little 
opposition, but O'Hara's men on the right 
encountered a considerable number of Jap- 
anese still alive and firing. 79 By 1 540 Com- 
pany F had recaptured two of the Marine 
batteries, the first one with the help of some 
of the Marine artillerymen who had re- 
mained in the vicinity after being driven 
off their guns. 80 A short while later the 
1 st Battalion reported that it had recap- 
tured the abandoned guns of Battery H of 
the 3d Battalion, 10th Marines. 81 By 1600 
the 106th Infantry was still 200 to 300 
yards short of positions of the 105th In- 
fantry, which it was supposed to relieve. 
Colonel Stebbins nevertheless decided to 
dig in where he was, although against the 
advice of one of his battalion commanders. 
Stebbins was concerned lest there be too 
many bypassed enemy to his rear. 82 This 
left the division commander with no alter- 
native but to evacuate the remainder of his 
isolated troops by water. 

Meanwhile, on the division right the 
165th Infantry was touched only lightly by 

72 For this day's action, the 105th Infantry (less 
3d Battalion and Company G) was awarded the 
Distinguished Unit Citation (Department of the 
Army, GO 49, 14 July 1948). 

73 lorith RCT Jnl, 7 Jul 44, Msg 1072. 

71 a 7th Inf Div G-3 Jnl, 7 Jul 44, Msgs 98, 37. 
73 Ibid,, Msg 39. 

7,i George W. Griner, Certificate, 12 Jul 44, 
Buc:kner Board Rpt, Exhibit ZZ. 

77 37th Inf Div G-3 Jnl, 7 Jul 44, Msg 46. 

TX I hid,, Msg 33. 

7!) 106th RCT Opns Rpt Foraokr, p. 23. 

8 " 1 06th RCT Jnl, 7 Jul 44, Msgs 1080, 1107. 

M1 Ibid., Msg 1 108. 

82 George W. Griner, Certificate, 12 Jul 44, 
Buckncr Board Rpt, Exhibit ZZ. 




7-9 July 1944 

• •• FffONT LINE ,EV£Nli*5 £ JOl' 

■ ■ Positions *eac*eq 8 Jul/ 

-^— PO|iT.0N5 REACHED 9 Jul* 
Cgnlatir >nt8r*dl tQQ fear 
ioao a iooo rn^oa 

f U s'i S 1 !- * 1 ^ I 

Marpt Point 

MAP 14 

the overflow from the charge that had all 
but overwhelmed the 105th. By 0930 the 
attached 3d Battalion, 106th, had finally 
reached the floor of Harakiri Gulch and 
was mopping up the Japanese still hidden 
in the caves and ditches. Occasionally, 
random riflemen who were apparently part 
of the main enemy counterattack wandered 
into the area to delay operations, but no 
serious opposition remained. Shortly after 
noon, the 1st Battalion, 165th, to the right, 
was able to advance through the draw at 
the upper end of the gulch and move into 

the plateau to the north overlooking the 
plain. It stopped there for the night and 
made plans to descend the cliffs the next 
day. 83 

Final Victory 

In spite of the fact that by nightfall of 
7 July the 27th Division had recovered 
some of the lost ground in the area of the 

HS iGf,th RGT Jnl, 7 Jul 44, Msgs 14, 15, 46; 
Love, Battle for Saipan, pp. 936-37. 



Flame Thrower Blasting Paradise Valley Cave,, an effective method of 
flushing Japanese, 

counterattack and had at last cleaned out 
Harakiri Gulch, General Holland Smith 
decided to relieve most of the Army units 
from the line. The 2d Marine Division 
(less detachments ) N4 was ordered to pass 
through the 27th Division and "mop up 
and destroy enemy elements" remaining in 
its zone of action. Attached to the Marine 
division would be the 165th Infantry, as 
well as the 3d Battalion, 6th Marines, 
which was now to be released from the 
control of the Army division. Upon being 

84 The detachments were the 2d Marines; ist 
Battalion, 29th Marines; and Company A, 2d 
Tank Battalion. 

passed through the 27th was ordered into 
corps reserve. 85 J^Map 7 j)| 

"Mop-up" was the proper word for the 
activities of most of the 2d Marine Division 
for the next two days. Along the beach 
the 6th Marines had an easy time after its 
jump-off at 1 1 30. In the words of the of- 
ficial report, "Initially, the attack was field 
day for our troops and slaughter for the 
Japs." Later in the afternoon, however, the 
regiment came up against a small pocket 
of resistance, just southwest of the coconut 
grove, containing about a hundred Japa- 

8i NTLF Opn Order, 7 Jul 44, NTLF Rpt Mar- 
ianas, Phase I, Incl A. 



ncse who had taken refuge in the bed of a 
small stream that ran down to the sea. 
Flame throwers, tanks, and self-propelled 
mounts were brought up to wipe out the 
enemy, but at nightfall the pocket still re- 
mained. By 1830 all units of the regiment 
had reached the beach, and the next day 
was spent mopping up and eliminating all 
enemy in the area. 8 ' 1 

In the hills and ravines just east of the 
coastal plain, the 8th Marines spent most 
of 8 and 9 July in demolition work, since 
the remnants of enemy consisted of disor- 
ganized groups holed up in caves. Every 
cave had to be investigated, and the few 
Japanese remaining alive were destroyed or 
driven out with hand grenades, flame 
throwers, and TNT charges. HT 

On the division right, the 165th Infan- 
try met similar scattered opposition and 
dealt with it in much the same way. De- 
laying its assault on the 8th until the 
marines could come abreast, the 165th 
jumped off from its positions north of 
Harakiri Gulch at 1130. By midaftcrnoon 
Company I had forced its way through 
Paradise Valley, where General Saito had 
established his final headquarters and from 
which the banzai order had been issued. 
Apparently not all of Saito's men had 
joined in the charge, for the caves in the 
cliffs' sides still harbored enough Japanese 
to offer stiff resistance to progress through 
the valley. By 1245 of g July forward ele- 
ments of the regiment reached their desti- 
nation on the shore, while Company K 
stayed behind to finish mopping up the 
caves of Paradise Valley. 88 

The morning of the 8th found the 4th 
Marine Division poised and ready for its 

final drive to Saipan's northern tip. On 
the left, the 23d Marines were on the 
high ground overlooking Karaberra Pass, 
through which the regiment would have to 
advance in order to seize its assigned por- 
tion of the shore line north of Makunsha. 
Following an intense preparation by rock- 
ets and tank fire, and assisted by LCI gun- 
boats lying off Makunsha, the 23d Marines 
forced its way through the pass and by 
1205 had rushed across the coastal flats to 
the sca. so On its right, the 23d and 24th 
Marines kept abreast and secured their as- 
signed zones by about 1530, while the 25th 
Marines, on the east flank, advanced its 
lines about 600 yards against no opposi- 
tion. That night the 1st Battalion, 2d 
Marines, suffered a scries of minor counter- 
attacks that were troublesome in the ex- 
treme but that failed to make a dent in 
the division line. On the gth, with three 
regiments abreast (the 23d Marines was 
pinched out), the 4th Marine Division 
completed the final lap in a fast sprint to 
Marpi Point. 90 At 16 15, 9 July 1944, Ad- 
miral Turner announced Saipan to be 
"secured." 91 

All that remained was to ferret out the 
few remaining enemy troops from their 
scattered hiding places in the caves and 
gulleys and ravines that Uttered the north- 
ern part of the island. Initially, this task 
was assigned to the two Marine divisions, 
with the 165th Infantry still attached to 
the 4th. These men still had to witness a 
few horrendous sights before they were 
through with Saipan. In spite of continu- 

"" 6th RCT SAR Forager, pp. 15-16. 

87 8th RCT SAR Foraoer, p. 9. 

HH 165th RCT Record of Opns Foragf.r, p. 9. 

so 4th Marine Div Opns Rpt Saipan, Annex H 
23d RCT Rpt, p. 48; Hoffman, Saipan, p. 238. 

'"' 4th Marine Div Opns Rpt Saipan, Sec. VI, 
pp. 36-37; Hoffman, Saipan, pp. 239-42. 

9) " NTLF Rpt Marianas, Phase' I, Incl E, G-3 
Rpt, G-3 Periodic Rpt 25. 



otis American efforts to induce both 
military and civilian survivors to give 
themselves up, the traditional Japanese 
code of death before surrender prevailed 
in most cases. Shortly after the declaration 
that the island was secured, hundreds of 
civilians leapt from the cliffs of Marpi 
Point to the knifelike rocks below. At times 
the waters below the point were so thick 
with the floating bodies of men, women, 
and children that naval small craft were 
unable to steer a course without running 
over them. HL ' 

On the 9th many Japanese soldiers swam 
out to the reefs of Tanapag Harbor and 
defied capture, 1st Lt. Kenneth J. Hensley, 
USMC, commanding officer of Company 
G, 6th Marines, was ordered out with a 
small flotilla of amphibian tractors to cap- 
ture or destroy these die-hards. A few sur- 
rendered, but most refused to give up. 
From one reef, to which fifty to sixty Jap- 
anese were clinging, machine guns opened 
up on the approaching LVT's. The Amer- 
icans returned fire and the force was an- 
nihilated. On another reef a Japanese 
officer was seen beheading his little band of 
enlisted men with his sword before he 
himself was shot down by his would-be 
captors. 93 

For the remainder of their brief stay on 
Saipan, the marines spent most of their 
days investigating the caves and wooded 
sections along the north shore. On 13 July 
the 3d Battalion, 6th Marines, occupied 
tiny Maniagassa Island in Tanapag Harbor 
in a miniature amphibious landing com- 

plete with naval gunfire, artillery, and 
aerial bombardment. They found twenty- 
nine Japanese soldiers on the island, but 
encountered no serious opposition. 04 This 
action brought to an end Marine activities 
on Saipan. The ad and 4th Divisions both 
withdrew to prepare for the assault on 
nearby Tinian. The Army took over the 
job of clearing out the last remnants of the 
enemy. From 31 July through 6 August, 
the 27th Division conducted a gradual 
sweeping operation with two regiments 
abreast from just north of Mount Tapot- 
chau to Marpi Point, thus concluding the 
organized mop-up. U5 Starting the middle of 
August and ending in October, the division 
embarked in stages for the much welcomed 
trip to the New Hebrides for rehabili- 
tation. 96 

The toll of American killed and wounded 
was high. Of the 71,034 officers and men 
that made up Holland Smith's Northern 
Troops and Landing Force, 87 it is esti- 
mated that 3,674 Army and 10,437 Ma- 
rine Corps personnel were killed, wounded, 
or missing in action. 98 This total of 14,1 1 1 
represents about 20 percent of the combat 
troops committed, or roughly the same 
percentage of casualties suffered at Tarawa 
and Peleliu, both of bloody renown," 

In exchange, almost the entire Japanese 

* 2 For eyewitness accounts of this episode, see 
Robert Sherrod, On To Westward: War in the 
Central Pacific (New York, Duell, Sloan and 
Pearce, Inc., 1945). 

ns NTLF Rpt Marianas, Phase I, Incl D, G-a 
Rpt, App. K. 

1,4 6th RCT SAR Forager, p. 17. 

!,s 37th Inf Div G-3 Rpt, G-3 Periodic Rpts 


!MJ 27th Inf Div, Hist, of 27th Inf Div, 20 Mar 

45. V- 4- 

97 Figure is from TF 51 Opns Rpt Marianas, 

p. 6. 

us Army casualty figures are derived from 27th 
Inf Div G-i Periodic Rpt, 6 Aug 44, Annex R, 
and XXIV Corps Final Rpt, S-i Rpt. Marine 
Corps figures were compiled by Machine Records 
Sec, Hq USMC, and published in Hoffman, Sai- 
pan, pp. 268-69. 

SB Smith, Approach to the Philippines, pp. 472- 
577; Stockman, The Battle for Tarawa, p. 67. 


garrison of about 30,000 men was wiped cracked, and American forces were at last 
out. Far more important, the inner defense within bombing range of the enemy home- 
line of the Japanese Kmpirc had been land. 




American and Japanese Plans 

Writing after the war, Admiral Spruancc 
expressed the opinion, "The Tinian opera- 
tion was probably the most brilliantly 
conceived and executed amphibious opera- 
tion in World War II." 1 To General 
Holland Smith's mind, "Tinian was the 
perfect amphibious operation in the Pacific 
war." 2 Historians have by and large — 
endorsed these sentiments. 3 

Much of this praise is well deserved, al- 
though a close examination of the facts re- 
veals that these, like most superlatives, are 
somewhat misleading. The invasion of 
Tinian, like other military operations, was 
not entirely without flaw. Various de- 
ficiences can be charged to both plan and 
execution. Yet, as an exercise in amphibi- 
ous skill it must be given a superior rating, 
and as a demonstration of ingenuity it 
stands as second to no other landing oper- 
ation in the Pacific war. 

Situated only about 3.5 miles off the 
southern coast of Saipan . Tinian i s the 
smaller of the two islands. \(Map lV.)\ YTova 
Ushi Point in the north to Lalo Point in 
the south, it measures about 12.5 miles, 
and in width it never extends much more 

1 Quoted in Maj. Carl W. Hoffman, USMC, 
The Seizure of Tinian, Historical Division, Head- 
quarters, U.S. Marine Corps (Washington, 1 95 1 ) 
(hereafter cited as Hoffman, Tinian), p. Ill, 

2 Smith, Coral and Brass, p. 201. 

3 See Isely and Crowl, U.S. Marines and Am- 
phibious War, p. 352; Morison, New Guinea and 
the Marianas, p. 351; Hoffman, Tinian, passim. 

than 5 miles.' 1 In one respect its terrain is 
not as formidable for would-be attackers 
than that of Saipan — it is far less moun- 
tainous. In the northern part of the island 
Mount Lasso rises to 564 feet, or only a 
little more than a third of the height of 
Tapotchau. Another hill mass of almost the 
same height dominates the southern tip of 
the island and terminates in heavily fis- 
sured cliffs that drop steeply into the sea. 
Most of the rest of the island is an undulat- 
ing plain, which in the summer of 1944 
was planted in neat checkerboard fields of 
sugar cane. 

It was indeed the relative flatness of 
Tinian's terrain that made it such a desir- 
able objective — that and the fact that its 
proximity to Saipan made its retention by 
the Japanese militarily inadmissible. Tin- 
ian's sweeping plains and gentle slopes 
offered better sites for bomber fields than 
its more mountainous sister island, and of 
course one of the main objectives of the 
Marianas operation was to obtain sites for 
air bases for very long range bombers. To 
a limited extent, the Japanese had already 
realized this possibility and near Ushi Point 
had constructed an airfield that boasted a 
runway almost a thousand feet longer than 
Aslito's. In addition, smaller fields were lo- 
cated just south of the Ushi Point field and 
at Gurguan Point, and another was under 

JICPOA Bull 73-74, 10 May 44, p. 2ofT. 



Coastal Area, Northwest Ttntan, showing White Beach areas, checkerboard 
terrain inland, and Ushi Pomt airfield in background, 

construction just northeast of Tinian 
Town. 5 

But if the island was well suited by na- 
ture for the construction of airfields, its 
natural features were also well disposed to 
obstruct a landing from the sea. Tinian is 
really a plateau jutting up from the sur- 
rounding ocean, and most of its coast line 
consists of cliffs rising sharply out of the 
water. Only in four places is this solid cliff 
line interrupted. Inland of Sunharon Bay, 
in the area of Tinian Town on the south- 
west coast, the land runs gradually to the 
sea, offering a fairly wide expanse of beach 
protected by the usual reef line. South of 
Asiga Point on the cast coast there is an 
indenture in the cliff wall that forms a 
small approachable beach about 125 yards 
in length. The northwest coast line offers 

other possible routes of ingress through the 
cliffs over two tiny beaches about 60 and 
1 50 yards in length, respectively. 6 

The peculiar features of the coast line 
placed American planners in a dilemma. 
The beaches off Tinian Town were obvi- 
ously the best suited for a landing opera- 
tion, but by the same token they were the 
best fortified and defended. The other 
beaches, which were little more than dents 
in the cliff line, were obviously not desir- 
able sites for an amphibious assaidt of 
corps dimensions. The risks of troops and 
supplies being congested to the point of im- 
mobility as they tried to pour through 
these narrow bottlenecks were considerable 
and alarming. For these reasons, which 
were just as apparent to the Japanese as 
to the Americans, defenses on the smaller 

5 NTLF Rpt Marianas, Phase III, Incl D, G-2 
Rpt, p. 32. 

6 Ibid., Annex A. 



beaches were less formidable than those 

In the end, the American planners 
seized the second rather than the first horn 
of their dilemma, chose the narrow beaches 
on the northwest coast, and accepted the 
risks that troops, equipment, and supplies 
might pile up in hopeless confusion at the 
water's edge. Having made the choice, the 
planners were compelled to devise special 
means of overcoming the accepted risks. 
This involved working out novel techniques 
that were radical modifications of standard 
amphibious doctrine as it had been evolved 
during the war in the Pacific. Paradoxi- 
cally then the invasion of Tinian was a 
"perfect amphibious operation" largely be- 
cause it was atypical rather than typical — 
because of its numerous departures from, 
rather than its strict adherence to, ac- 
cepted amphibious doctrine. 

Plan for the Invasion 

From the very outset of the planning for 
the seizure of the southern Marianas, Tin- 
ian had been considered one of the three 
main targets of the operation. Holland 
Smith's Northern Troops and Landing 
Force was ordered to "land, seize, occupy 
and defend SAIPAN Island, and then 
, , . be prepared for further operations to 
seize, occupy and defend TINIAN Is- 
land," 7 Consequently, planning for the 
Tinian phase commenced at the same time 
as that for the capture of Saipan and was 
continuous until the very day of the land- 
ing on Saipan, By the time Admiral 
Turner's task force set sail from Pearl Har- 
bor, maps, photographs, and charts of 
Tinian had been distributed and tentative 

arrangements had been made for loading 
and for resupply shipping. While at sea, 
Holland Smith's staff had more leisure than 
earlier to concern itself with this phase of 
the operation, and by the time the ships 
reached Eniwetok a draft plan was ready 
for the commanding general. In devising 
this plan, the staff gave due consideration 
to the relative merits of the various landing 
beaches and recommended that a landing 
be made on northern Tinian in order to 
make full use of artillery emplaccd on 
southern Saipan. 8 

While the fighting for Saipan was in 
process, the Americans were afforded ideal 
opportunities for scrutinizing the island to 
the south from every angle. Beginning on 
20 June, when artillery first bombarded 
Tinian from southern Saipan, 9 observation 
planes flew daily over northern Tinian. 
Frequent photo reconnaissance missions 
were flown, and many valuable documents 
throwing light on Tinian's defenses were 
captured on Saipan, 10 Opportunities for 
gathering intelligence were almost without 
limit, and it is doubtful if any single en- 
emy island was better reconnoitcred during 
the Pacific war. 

With Saipan secured and the prepara- 
tions for the next landing in mid-passage, 
a change in command within the Northern 
Troops and Landing Force was ordered. 
On 12 July General Holland Smith was 
relieved and ordered to take command of 
Fleet Marine Force, Pacific, a newly cre- 
ated headquarters for all Marine Corps 
combat units in the theater. The new com- 
manding general of Northern Troops and 
Landing Force was General Schmidt, who 

7 NTLF Opn Plan 30-44, 10 May 44, in 
NTLF Rpt Marianas, Phase I, Incl A. 

s NTLF Rpt Marianas, Phase III, p, 3. 
XXIV Corps Arty Final Rpt on Foraoer 
Opn, S-3 Rpt, p. 6. 

10 NTLF Rpt Marianas, Phase III, p. 3. 



was in turn relieved of his command of the 
4th Marine Division by General Gates. 11 
Concurrently, a shift in the naval com- 
mand structure took place. Admiral Hill, 
who had served as Admiral Turner's sec- 
ond in command, took over a reconstituted 
Northern Attack Force (Task Force 52) 
and thus became responsible, under the 
Commander, Joint Expeditionary Force 
(Admiral Turner as Commander, Task 
Force 51), for the capture of Tinian. 13 

As planning for Tinian went into high 
gear, it was becoming increasingly appar- 
ent to all hands that the original concept 
of landing the assault troops somewhere 
in the northern part of the island was 
sound. Members of the staff of the 4th Ma- 
rine Division, notably Lt. Col. Evans F. 
Carlson, the division's planning officer, had 
already decided that an amphibious land- 
ing in this area was desirable. Working 
independently of the Marines, Admiral Hill 
had arrived at the same conclusion. 13 All 
agreed that the Tinian Town area was too 
well defended to justify an amphibious as- 
sault there and that the advantages of 
heavy artillery support for landings on the 
northern beaches were too considerable to 

All, that is, but one. Admiral Turner 
was still not convinced, In his mind, the 
Tinian Town beaches offered important 
advantages that should not be lightly dis- 
missed. From the point of view of gradient 
and inland approaches, the Tinian Town 
beaches were even more favorable to the 
attacker than those used on Saipan and 
certainly far better than Tinian's other 
beaches. Also, Sunharon Bay offered an ex- 
cellent protected harbor for small craft and 

11 Ibid,, p. 4; Smith, Coral and Brass, p. Qoi. 

12 TF 51 Opn Rpt Marianas, p. 1. 

13 Hoffman, Tinian, pp. 20, iGr. 

good facilities for unloading supplies, once 
the beachhead was secured. On the other 
hand, the beaches in the northern half of 
the island, argued the admiral, were too 
narrow to permit a rapid landing of a force 
of two divisions with full supplies and 
equipment, and if the weather took a turn 
for the worse the shore-to-shore movement 
of supplies in small craft from Saipan 
might be seriously endangered. In addition, 
an advance down the full length of the 
island would take too much time, and the 
troops would soon outrun their artillery 
support based on Saipan — an especially 
dangerous prospect should weather condi- 
tions forbid shifting the heavy artillery 
pieces from Saipan to Tinian. 14 

In the light of these objections and out 
of ordinary considerations of military cau- 
tion, General Schmidt ordered a physical 
reconnaissance of the northern beaches. 
The task fell to the Amphibious Reconnais- 
sance Battalion, V Amphibious Corps, 
commanded by Capt. James L. Jones, 
USMCR, and naval Underwater Demoli- 
tion Teams 5 and 7, commanded by Lt. 
Comdr, Draper L. KaufTman, USN, and 
Lt. Richard F. Burke, USN, respectively. 
Their job was to reconnoiter Yellow Beach 
1 on the eastern coast below Asiga Point 
and White Beaches 1 and 2 on the north- 
western coast. Under cover of darkness the 
three groups were to be carried part way 
to their destinations by the high-speed 
transports Gilmer and Stringham. Then, 
launched in rubber landing boats (LCR's), 
they would be paddled to distances about 
500 yards offshore and swim in the rest of 
the way. The men were charged with the 
responsibility of investigating and securing 

1 4 Ltr, Turner to Comdt USMC, 2 1 Dec f,o, 
quoted in Hoffman, Tinian, p. 2 1 . 



Marianas Leaders Confer at Tinian. Left to right: Rear Adm. Harry Hill, 
Maj. Gen. Harry A. Schmidt, Admiral Spruance, General Holland Smith, Admiral 
Turner, Maj, Gen, Thomas E. Watson, and Maj, Gen. Clifton B, Gates. 

accurate information concerning the height 
of surf, the height and nature of the reef 
shelf, depth of water, location and nature 
of mines and underwater obstacles, the 
slope of the bottom off the beaches, the 
height and nature of cliffs flanking and 
behind the beaches, exits for vehicles, 
and the nature of vegetation behind the 
beaches. The naval personnel would 
conduct the hydrographic reconnaissance 
while members of the Marine amphibious 
reconnaissance group were to reconnoiter 
the beaches themselves and the terrain 
inland. 1,5 

After dark on 10 July, but well before 
moonrise, Gilmer and Stringham got under 

lr NTLF Rpt Marianas, Phase III, Incl L, 
Amph Ren Bn Rpt, Annex A, pp. 1-3. 

way from Magiciennc Bay on the east coast 
of Saipan to take their respective stations 
off of Yellow Beach 1 and White Beaches 
1 and 2. As the rubber boats approached 
Yellow Beach 1, the men heard sharp re- 
ports and thought they were being fired on, 
but went about their business anyway. 
Two of the men swam along the cliffs south 
of the beach and discovered them to be 
20 to 25 feet high and unscalable by in- 
fantry without ladders or nets. One Marine 
officer, 2d Lt. Donald Neff, left two of his 
men at the highwater mark and worked 
his way along inland for some thirty yards 
to investigate the possibilities for vehicle 
exits. Japanese sentries were apparently 
patrolling the entire area, but the sus- 
pected rifle shots proved to be exploding 
caps being used by construction workers 



nearby. In any case, all hands got back to 
their ships without being detected. 11 ' 

Meanwhile, on the other side of the is- 
land, the reconnaissance of White Beaches 
i and 2 hit a snag. As the rubber boats 
cast off they were set rapidly to the north 
by a strong current that they had not been 
compensated for. Hence the swimmers as- 
signed to White Beach 2, the southernmost 
of the two, ended up on White 1, while 
the second group destined for the latter 
beach were set ashore about 800 yards to 
the north. This left White 2 unrecon- 
noitered, and next night another group of 
swimmers had to return to finish the job. 17 

The information gathered during the 
two nights fully justified the valiant labor 
expended. Yellow Beach 1 was clearly un- 
suitable for an amphibious landing. In ad- 
dition to its natural disadvantages, the 
beach was strung with strong double-apron 
wire, and large, floating, contact mines 
were found anchored about a foot under- 
water off the reef. 18 On the other side of 
the island no man-made obstacles were re- 
ported on White Beaches 1 and 2. Al- 
though White Beach 1 to the north was 
only sixty yards in length, the bluffs that 
flanked it for about 150 yards on either 
side were only from six to ten feet in height 
and offered enough small breaks to permit 
men to proceed inland in single file without 
the need of cargo nets or scaling ladders. 
From the reef to the shore line the water 
depth was never more than four feet and 
the gradient was slight. Of the hundred 
and fifty yards of White Beach 2, only the, 
central seventy yards were approachable by 
amphibian vehicles, the flanks of the beach 
being guarded by coral barriers jutting out 

from the reef. Nevertheless, the barriers of- 
fered no obstacle to infantrymen, who 
could scramble over them and wade the 
rest of the way in. At two hours before 
high tide the water inside the reef was no- 
where more than four feet in depth. 19 In 
short, although the White Beaches were far 
from ideal for landing purposes, they were 
better than Yellow Beach 1 , and except for 
their narrowness offered no known natural 
or man-made obstacles. 

With this information in hand, Admiral 
Turner's objections to a landing on the 
northwest coast, however strong they may 
once have been, were overcome. At a 
meeting held aboard his flagship on 12 
July, General Schmidt made a forceful 
presentation of the case for the White 
Beaches. An amphibious assault against the 
strong enemy defenses in the Tinian Town 
area would be too costly; artillery could 
be more profitably employed against the 
northern beaches; Ushi Point airfield 
would be more quickly seized and made 
ready; tactical surprise could be obtained; 
the operation could more easily be con- 
ducted as a shorc-to-shore movement from 
Saipan; and, finally, most of the supplies 
could be preloaded on Saipan and moved 
on amphibian tractors and trucks directly 
to inland dumps on Tinian. Admiral Hill 
concurred, and Admirals Turner and 
Spruance gave their consent to a landing 
on White Beaches 1 and 2. 2 " 

The next day, 13 July, General Schmidt 
issued the operation plan that was to gov- 
ern the invasion of Tinian. 21 General 
Gates' 4th Marine Division was assigned 
the task of conducting the amphibious as- 

1 '' Ibid., Annex A; Annex E. 
' 7 Ibid., Annex B. 
18 Ibid., Annex E, 

111 Ibid. , Annexes D and E. 

20 Hoffman, Tinian, pp. 23-24. 

21 NTLF Opn Plan 30-44, 13 Jul 44. NTLF 
Rpt Marianas, Phase III, Incl A. 



sault over White Beaches i and 2 on JIG 
Day, which was later established as 24 
July. On landing, the division was to make 
its main effort toward Mount Lasso and, 
before reorganizing, seize the Force Beach- 
head Line, which included Faibus San Hilo 
Point, Mount Lasso, and Asiga Point. 
Once this area was captured, it was pre- 
sumed that the beachhead would be safe 
from ground-observed enemy artillery fire. 
To accomplish the division's mission Gen- 
eral Cates ordered the 24th Marines to 
land in column of battalions on White 
Beach 1 on the left, the 25th Marines with 
two battalions abreast on White Beach 2. 
The 23d Marines would remain in division 
reserve. 22 

The assault troops would be carried 
ashore in the customary fashion in am- 
phibian tractors discharged fully loaded 
from LST's. Of the 415 tractors assigned 
to carry troops, 225 were supplied by 
Army units — the 715th, 773d, and 534th 
Amphibian Tractor Battalions. The re- 
mainder were Marine LVT's from the 2d, 
5th, and 10th Amphibian Tractor Battal- 
ions. Because of the narrowness of the land- 
ing beaches, only one company of amphib- 
ian tanks could be employed in the assault, 
Company D, 2d Armored Amphibian 
Tractor Battalion (Marine), The battalion 
was ordered to precede the first wave of 
troops, fire on the beaches after naval gun- 
fire was lifted, and move to the flanks 
before reaching land. The 708th Armored 
Amphibian Tank Battalion (Army) was 
ordered to stand by off the beaches and 
be prepared to land and support the in- 
fantry ashore. 23 

As before, command of the entire opera- 
tion was vested in Admiral Turner as 
Commander, Joint Expeditionary Force 
(Task Force 51), under Admiral Spru- 
ance; General Holland Smith, who still 
retained his position of Commander, Ex- 
peditionary Troops, continued in over-all 
command of troops ashore. In fact, how- 
ever, both of these officers had sailed 
aboard Rocky Mount on 20 July to be on 
hand for the Guam landings, which took 
place the next day, and did not return to 
the Saipan-Tinian area until the 25th. 24 
During the landing then, Admiral Hill, as 
Commander, Northern Attack Force (Task 
Force 52), commanded all naval craft and 
supporting forces, while General Schmidt 
commanded the landing forces. 23 Even 
after Admiral Turner returned, Admiral 
Hill retained the responsibility "for offen- 
sive and defensive surface and air action" 
in the area and for all practical purposes 
Schmidt remained in tactical control of the 
troops. 26 

Because most of the heavy artillery 
pieces could more profitably be employed 
from emplacements on Saipan, the 4th Ma- 
rine Division would carry only 75-mm. 
pack howitzers in the initial assault. In ad- 
dition to its own two battalions ( 1 st 
and 2d Battalions, 14th Marines), it was 
loaned the two light battalions of the 2d 
Marine Division (1st and 2d Battalions, 
10th Marines). These battalions would be 
carried ashore in Marine DUKWs. Addi- 
tional fire power was afforded the division 
by attaching the 2d Division's tank bat- 
talion. Army troops (1341st Engineer 

23 4th Marine Div Opn Order 34-44. 
23 NTLF Rpt Marianas, Phase III, Incl H, 
LVT Rpt, pp. 1-3. 

24 TF 51 Opn Rpt MarianaSj Incl A, pp. 

25 NTLF Opn Plan 30-44, Annex I. 

afi TF 51 Opn Rpt Marianas, Incl A, pp. 
r 7-'9. 



LVT With Ramp 

Battalion) would make up part of the as- 
sault division's shore party, the remainder 
being provided by the 2d Battalion, 20th 
Marines. 27 

To the rest of General Watson's 2d Ma- 
rine Division was assigned the role of 
landing in the rear of the assault division 
once the latter had cleared an initial beach- 
head and moved inland. Before this, the 
division was to conduct a demonstration 
landing off Tinian Town for the purpose of 
diverting Japanese attention from the 
main assault to the north, 28 

The 27th Infantry Division, less the 
105th Infantry and less its organic artil- 
lery, was to remain in corps reserve and 
"be prepared to embark in landing craft 

' 21 NTLF Opn Plan 30-44, 13 Jul 44. 

28 This scheme was not part of the original 
operation plan, but was devised shortly before 
the assault took place. 

on 4 hours notice and land on order , , . 
on Tinian." 

One of the main justifications for the 
final decision to land over the unlikely 
beaches on the northwestern shore of the 
island was the feasibility of full exploita- 
tion of artillery firing from Saipan. Con- 
sequently, all of the field pieces in the area 
except for the four battalions of 75-mm. 
pack howitzers were turned over to XXIV 
Corps Artillery during the preliminary and 
landing phase. General Harper arranged 
his thirteen battalions, totaling 156 guns 
and howitzers, into three groupments, all 
cmplaced on southern Saipan. Groupment 
A, commanded by Col. Raphael Griffin, 
USMC, consisted of five 105-mm. battal- 
ions, two each from the Marine divisions 
and one from V Amphibious Corps. It was 
to reinforce the fires of the 75-mm. pack 
howitzers and be ready to move to Tinian 



on order. Groupment B, under the 27th 
Division's artillery commander, General 
Kernan, was made up of all of that divi- 
sion's organic artillery except the 106th 
Field Artillery Battalion. It was to reinforce, 
the fires of Groupment A and also to be 
ready to displace to Tinian. Groupment C, 
commanded by General Harper himself, 
contained all the howitzers and guns of 
XXIV Corps Artillery plus the 106th Field 
Artillery Battalion. It was to support the 
attack with counterbattery, neutralization, 
and harassing fire before the day of the 
landing, deliver a half-hour preparation on 
the landing beaches immediately before the 
scheduled touchdown, and execute long- 
range counterbattery, harassing, and inter- 
diction fire. 29 

In addition to the artillery, the troops 
would of course have the support of 
carrier-borne aircraft, aircraft based on 
Aslito field, and naval gunfire. Although 
all three supporting arms were to be cm- 
ployed against targets everywhere on Tin- 
ian, primary responsibility for the northern 
half was allocated to artillery while naval 
gunfire and air took over the southern half. 
The task of co-ordinating the three was 
vested in a XXIV Corps Artillery repre- 
sentative at General Schmidts head- 
quarters. 30 

The most unique feature of the plan for 
Tinian was its logistical provisions. Because 
only slightly more than 200 yards of beach 
were available, it was essential that precau- 
tions be taken to avoid congestion. Hence, 
a supply plan was developed that allowed 
all supplies to cross the beach on wheels or 
tracks and move directly to division dumps 
without rchandling. This entailed devising 

a double shuttle system in which loaded 
trucks and Athey trailers traveled back and 
forth between the base supply dumps on 
Saipan and division supply dumps on Tin- 
ian, and all amphibian vehicles carrying 
supplies between ship and shore moved di- 
rectly to division dumps. The objective was 
to avoid any manhandling of supplies on 
the beaches themselves. The solution rep- 
resented a marked departure from standard 
amphibious practice and was made possi- 
ble, of course, by the proximity of Tinian 
to the supply center on Saipan. 31 

The plan called for preloading thirty-two 
LST's and two LSD's at Saipan with top- 
deck loads of all necessary supplies except 
fuel. Ten LST's were allotted to each Ma- 
rine division, eight to general reserve, and 
four primarily to 75-mm. artillery. All am- 
phibian tractors and trucks available, both 
Army and Marine Corps, were initially as- 
signed to the 4th Marine Division, but after 
the assault was over were to be distributed 
between the two divisions. The supplies 
were loaded on the LST's in slings, and 
the ships carried crawler cranes on their 
top decks so that the slings could drop sup- 
plies into DUKW's and LVT's coming 
alongside. To carry out the shuttle system, 
the plan called for preloading eighty-eight 
cargo trucks and twenty-five Athey trail- 
ers on Saipan to be taken to Tinian aboard 
LCT's and LCM's. A special provision for 
fuel supply was made. Seven ponton barges 
loaded with drums of captured Japanese 
gasoline and matching lubricants were to 
be towed to positions off the landing 
beaches to act as floating supply and fuel- 
ing points for LVT's and DUKW's. Other 
fuels for initially refueling the trucks were 

as NTLF Opn Plan 30-44, Annex F. 
ao Ibid., Annexes C, D, F. 

3X NTLF Rpt Marianas, Phase III, Incl F, 
G-4 Rpt 3 p. 1. 



placed on barges that were to be spotted 
off the beaches. 32 

One other innovation introduced in the 
Tinian campaign was a special portable 
LVT bow ramp. Ten amphibian tractors 
were equipped with this device so as to 
provide a means for extending the narrow 
beach area. The ramps were so constructed 
that an LVT could drive up to a cliff 
flanking the beaches, place the ramp in 
position along the ledge, then back down 
leaving the ramp to act as a sort of cause- 
way by which other vehicles could get to 
shore. 33 

Finally, precautions were taken to sup- 
ply the troops in case of unexpected bad 
weather after the landing. Plans were 
made to drop about 30 tons of supplies by 
parachute and to deliver 100 tons by air 
daily to Ushi Point airfield as soon as it 
had been captured. 31 

The Enemy 

As already observed, the opportunities 
for gaining detailed intelligence of Tinian's 
defenses were superior to those enjoyed by 
American forces in most Pacific operations. 
Proof of this superiority lies in the accuracy 
with which General Schmidt's staff was 
able to estimate Japanese strength and dis- 
positions. As of 13 July they predicted, on 
the basis of captured documents, photo re- 
connaissance, and other intelligence data, 
that the strength of the Tinian garrison 
came to 8,350, plus possible home guard 
units. The main part of this force was be- 
lieved to consist of the 50th Infantry Reg- 
iment (reinforced) — about 4,000 men — 

32 ibid., p. 2. 

■ 1:l NTLF Rpt Marianas, Phase III, Incl H, 
LVT Rpt, p. 3 . 

:i4 TF 52 Rpt Tinian, pp. 1-3. 

and the 56th Keibitai (Naval Guard 
Force ) — about 1 , 1 00 men — plus sundry air 
defense, base force, and construction per- 
sonnel. The Army troops were believed to 
be disposed in three sectors, northern, 
western, and southern, which included re- 
spectively the Ushi Point-Asiga Bay re- 
tion, the west coast north of Gurguan 
Point including White Beaches 1 and 2, 
and the southern part of the island includ- 
ing Tinian Town. The northern and south- 
ern sectors were thought to be defended by 
at least one infantry battalion each, but 
the western sector where White Beaches 1 
and 2 were located had, it was estimated, 
only one company with one antitank 
squad. It was predicted that in each of 
these sectors the Japanese would first try 
to repulse the landing at the water's edge 
and would shift two thirds of each defense 
force from the areas not under attack to 
the beaches where the actual landings were 
taking place. A reserve force of one bat- 
talion was believed to be located near 
Mount Lasso, and it too was expected to 
move to the specific area under amphibious 
attack. One artillery battalion was thought 
to be located in the Tinian Harbor area, 
one battery near Asiga Bay. These esti- 
mates, except those pertaining to artillery 
strength, were remarkably accurate. 

The defense of Tinian was in the charge 
of Col. Takashi Ogata, commanding offi- 
cer, 50th Infantry Regiment, which repre- 
sented the bulk of the Japanese Army 
forces on the island. Other important units 
were the 1st Battalion, 135th Infantry; the 
Tank Company, 18th Infantry; the 56/A 
Naval Guard Force; and two naval anti- 
aircraft defense units. All together, Ogata 
had four Army infantry battalions, none of 
which were straggler units, plus additional 


Table 1 — Estimated Strength op the Japanese Garrison on Tinian 

Unit Unit Commander Strength 

Grand total 8,039 

Army — total 3,929 

50 tk Infantry Regiment Col. Takashi Ogata 

Headquarters — 60 

1st Battalion .... , . — a576 

2d Battalion — 576 

3d Battalion — 576 

Artillery Battalion (12 75-mm. mountain guns) Maj. Katuro Kahi 360 

Engineer Company Lt. Chukhi Yanu 169 

Antitank Platoon (6 37-mm. antitank guns) 2d Lt. Moto Otani 42 

Signal Company ,.Lt. Ilayashi 141 

Supply Company Lt. Kenishi Nozaki 200 

Medical Company. ... Lt. Masaakira Narazawa 130 

Fortification Detachment ... ... Capt. Masagi Hiruma 60 

1st Battalion, 135th Infantry Regiment Capt. Isumi B714 

18th Infantry Regiment Tank Company (9 tanks and 2 amphibian 

trucks),... Lt. Katsuo Shikamura c 65 

264th Independent. Motor Transport Company platoon — 60 

29th Field Hospital Detachment — 200 

Navy — total.. 4,110 

56th Naval Guard Force Capt. Goichi Oya 950 

82d Antiaircraft Defense Unit (24 25-mm. antitank guns) Lt,(s.g.)Kichitaro Tanaka 200 

83d Antiaircraft Defense Unit (6 dual-purpose 75-mm. guns) Lt.(s.g.)Meiki Tanaka 250 

233d Construction Unit... — 600 

Headquarters, 1st Air Fleet ..Vice Adm. Kakuji Kakuta 200 

Air units (mostly 523d TOKA) - 1,110 

Miscellaneous construction personnel ,. — 800 


aThe strength figures for the three infantry regiments given here arc somewhat lower than those estimated by NTLF, chiefly because 
the latter included attached artillery units in infantry strcuigt.h estimates. The figure 576 is the actual strength of the 2d Battalion, $utk 
Infantry, as of February 1944 (Headquarters, 2d Battalion, $Oth Infantry Regiment, War Diary, February 1944, NA 27434). It is assumed that 
the other regiments were approximately the same. 

hEstjmated on the basis of unit records for May 1944 of Headquarters, ist and 3d Companies, and Infantry Gun Company (NA 22237, 
27394, 27393). 

eShikanwra Tai War Diary, 29 April to 23 July 1944 (NA 22831). 

Source: These strength figures are derived frorn NTI.F Rpt Marianas, Phase III, Incl A, G-2 Rpt, pp. 24-30 and TF 56 Rpt Forager, 
Annex A, G-2 Rpt, p. 57. 

infantry in the 56th Naval Guard Force units. The 18th Infantry Tank Company 
and other naval units. For artillery, the had nine tanks, which constituted the en- 
Japanese commander had his regimental tire armored strength present. Total pcr- 
artillery battalion, the coast artillery sonnel strength, as indicated in Table 1, 
manned by part of the 56th Naval Guard came to a little more than eight thousand 
Force, and two naval antiaircraft defense officers and men, Army and Navy. 



As foreseen by General Schmidt's intelli- 
gence section, the Japanese Army plan for 
the defense of Tinian provided for the dis- 
position of forces in three sectors. (Map 
\i5)\ The northern sector force guarding 
Ushi Point, Asiga Bay, and part of Masalog 
Point was the responsibility of the 2d Bat- 
talion, 50th Infantry, and a platoon of 
engineers; the western sector, containing 
Mount Lasso and the northwest coast, was 
guarded only by the 3d Company, 1st Bat- 
talion, 50th Infantry, and an antitank 
squad. Regimental reserve in the southern 
sector consisted of the 1st Battalion, 50th 
Infantry, less the 3d Company and less one 
antitank squad and was located about 
3,000 yards southeast of Faibus San Hilo 
Point. The 1st Battalion, 135th Infan- 
try, was designated "mobile counterattack 
force," and was in effect another reserve, 
Ogata's armored strength came to only 
nine tanks of the Tank Company, 18th In- 
fantry, which was located on the northeast 
side of Marpo Well with orders to advance 
either to Tinian Town or Asiga Bay, wher- 
ever the landings came. In addition, this 
company had two vehicles rarely found 
among Japanese forces, amphibious trucks 
similar to the American DUKW. 35 

Japanese naval personnel on the island 
were under the command of Capt. Goichi 
Oya, who reported to Colonel Ogata. 
There was another, more senior, naval offi- 
cer present on the island, but he held no 
position in the chain of command and had 
nothing to do with the defense of Tinian. 
This was Vice Adm. Kakuji Kakuta, Com- 
mander in Chief, 1st Air Fleet, who was 
responsible only to Admiral Nagumo of the 
Central Pacific Area Fleet. After the loss of 
most of his planes in the Battle of the 

Philippine Sea, Kakuta made several efforts 
to escape Tinian by submarine. Each time 
he failed, and in the end he apparently 
committed suicide. 36 

Captain Oya appears to have made some 
effort to integrate his command with that 
of the Army. The 56th Naval Guard Force 
was charged with the defense of the air 
bases, defense of harbor installations and 
ships in the harbor, and destruction of en- 
emy attack forces. The force was divided 
into two parts. One was to man the coastal 
defense guns and antiaircraft weapons and 
the other, called the Coastal Security 
Force, was to maintain small coastal patrol 
boats and lay beach mines. No matter what 
the intentions of cither commander, how- 
ever, it would seem that there was little 
real co-ordination or even co-operation be- 
tween Army and Navy forces. There may 
have been serious interservice friction. 37 
This is at least suggested in the captured 
diary of one Army noncommissioned offi- 
cer, who wrote : 

15 June: The Naval aviators arc robbers. 
There aren't any planes. When they ran off 
to the mountains, they stole Army provisions, 
robbed people of their fruits and took ears, 

25 June: Sailors have stolen our provi- 
sions. They took food off to the mountains. 
We must bear with such until the day of de- 
cisive battle. . . . 

6 July: Did Vice-Admiral Kakuta when 
he heard that the enemy had entered our 
area go to sleep with joy? 38 

35 NTLF Rpt Marianas, Phase III, Incl A, 
G-2 Rpt, pp. 8-9. 

36 Ibid., p. ai; 4th Marine Div Opns Rpt Tin- 
ian, G-2 Rpt, Special Intel Rpt 4. 

37 4th Marine Div, Representative Translations 
Made on Tinian, Record and Research Sec, Hist 
Br, G-3 Hq USMC, Sec. I, Sec. IX, p. 3. 

SR CINCPAC-CINCPOA Trans 11405, ex- 
cerpts from the diary of a noncommissioned 
officer, a member of the Medical Administrative 
Unit, Mountain Artillery Battalion, 50th Infan- 
try Regiment. 

7 7-mmHvMG 

J&-mm Iwirt-mou^l 

In Airfield Area ! 

S-ii-mT AA/AT 

iFouf 20'tnm AAfljTcmati'C cannons 
fifteen £5-mm t¥*in*inount 
Si i 75-mm AA 

yo^i]'L.'g Po«** 

Gurgucm Point 

1 20-mm duo>purp05e 


25-rom twin-mount 

IZO-mm dual-purpose 
Mo*po Point 




t_oio Point 


sectoh bqunoah* 
Japanese unit 


Fixed sun position and 
number of suns 


, 1 ■ 


jooo va*ds 


MAP 15 



Responsibility for coastal defense was di- 
vided about equally between Army and 
Navy. Because of the small number of 
beaches over which hostile troops could 
possibly land, the problem was somewhat 
simplified. Consequently, even with the 
rather limited means at hand, it was pos- 
sible for the Japanese to distribute their 
fixed gun positions so as to place a fairly 
heavy guard around the only feasible ap- 
proaches to the shore. The Tinian Town 
area boasted three British-made 6-inch 
coastal defense guns, two 75-mm. moun- 
tain guns, 38 and six 25-mm, twin-mount 
antiaircraft and antitank automatic can- 
nons. Just up the coast from Tinian Town 
in the area of Gurguan Point were three 
1 20-mm. naval dual-purpose guns and nine 
25-mm, twin mounts that guarded the 
northern approaches to Sunharon Bay as 
well as the Gurguan Point airfield. The 
northwest coast from Ushi Point to Faibus 
San Hilo Point, including the area of White 
Beaches 1 and 2, was quite well fortified, 
especially considering that the Japanese 
had no real expectations of hostile am- 
phibious landings in that area. All to- 
gether, this stretch of coast line contained 
three 140-mm. coastal defense guns, two 
75-mm. mountain guns, two 7.7-mm. 
heavy machine guns in pillboxes, one 37- 
mm. covered antitank gun, two 13-mm. 
antiaircraft and antitank machine guns, 
two 76.2-mm. dual-purpose guns, and 
three 1 20-mm, naval and dual-purpose 
guns. In addition, in the hills behind and 
within range of this shore line were two 

:|U The JICPOA report cited below, note 45, 
lists these as 75-mm. M94 mountain howitzers. 
This must be an error because the Japanese had 
no 7f,-mm. howitzers and their Model 94 field 
piece was a 75-mm. mountain (pack) gun, (War 
Dept Technical Manual E 30-480, 15 Sep 44, 
Handbook on Japanese Military Forces, p. yao.) 

47-mm. antitank guns, one 37-mm. anti- 
tank gun, and five 75-mm. mountain 
guns. Guarding Ushi Point airfield were 
six 13-mm. antiaircraft and antitank guns, 
fifteen 25-mm, twin mounts, four 20-mm. 
antiaircraft automatic cannons, and six 
75-mm. antiaircraft guns. On the north- 
east coast, between Ushi Point and Masa- 
log Point, were seven 140-mm. coastal de- 
fense guns, three 76.2-mm. dual-purpose 
guns, one 37-mm. antitank gun, and 
twenty-three pillboxes containing machine 
guns of unknown caliber. Except for the 
coastal defense guns, all of these weapons 
were concentrated in the area of Yellow 
Beach 1, south of Asiga Point. Finally, in- 
land from Marpo Point on the southeast 
coast there were four 1 20-mm. dual- 
purpose guns. 40 

The most surprising feature of the dis- 
tribution of fixed positions is the relatively 
heavy concentration of guns within range 
of White Beaches 1 and 2. In spite of the 
fact that General Ogata assigned a low pri- 
ority to the infantry defenses in that re- 
gion, it is quite apparent that the Japa- 
nese were by no means entirely neglectful 
of the area. The figures cited here of course 
give no indication of the damage wrought 
on these positions by naval gunfire, field 
artillery, and aerial bombardment before 
the landing. But had American intelligence 
estimates of Japanese artillery dispositions 
been as accurate as they were in other re- 
spects, the plan for an amphibious landing 
over the White Beaches might not have 
been undertaken so optimistically. 

More accurate knowledge of Japanese 
mining activities off of White Beaches 1 
and 2 might also have given the American 

4,1 JICPOA Study of Japanese Fixed Gun Po- 
sitions, 24 Jul-f, Aug; Hoffman, Tinian, p. 15. 



planners pause. The reports of the amphib- 
ious reconnaissance and underwater demo- 
lition groups to the contrary, the Japanese 
had set up a mine defense of sorts along 
the northwest coast. Off White Beach i 
they had laid a dozen horned mines, though 
by the time of the landing these had de- 
teriorated to the point of impotence. White 
Beach 2 was mined in depth. Hemispher- 
ical mines were placed in two lines off- 
shore, conical yardstick, and box mines 
covered the exits from the beach. All to- 
gether, more than a hundred horned mines 
were laid in the area. In addition, there 
were many antipersonnel mines and booby 
traps concealed in cases of beer, watches, 
and souvenir items scattered inland. On 
the other side of the island, Yellow Beach 
1 was protected by twenty-three horned 
mines and by double-apron barbed wire. 
In the Tinian Town area a strip about 
thirty-five yards wide from the pier north 
along the water's edge to the sugar mill was 
completely mined. The beach south of the 
pier was laid with hemispherical mines that 
had steel rods lashed across the horns. Be- 
hind these were conical mines placed in 
natural lanes of approach from the shore 

Until the very eve of the landing, the 
Japanese worked furiously to improve their 
beach defenses, especially in the Tinian 
Town and Asiga Bay areas. Even the 
gathering rain of American shells and 
bombs failed to stop them entirely, for 
when the pressure became too great they 
worked at night and holed up during the 
day. 41 

41 CINCPAC-CINCPOA Trans 11962, ex- 
cerpts from the diary of Kancko Tokutaro, non- 
commissioned officer at Hq, sd Bn, 50th Inf 
Regt; Shikamura Tai War Diary, 29 Apr to 23 
Jul, NA 22O31. 

Ogata was well aware that an invasion 
of Tinian was inevitable, and in one respect 
he was more fortunate than Saito. 42 Unlike 
the commanding general of Saipan, he had 
no stragglers to deal with, and his Army 
troops were well trained, well equipped, 
and well integrated under a unified com- 
mand. He had had his regiment since 
August of 1940. For almost four years 
before moving to Tinian the unit had been 
stationed in Manchuria, and, under the 
semifield conditions obtaining there, Ogata 
was able to develop a high degree of homo- 
geneity and esprit.™ 

Ogata's plan of defense conformed to 
standard Japanese doctrine at this stage 
in the war. The enemy was to be destroyed 
at the water's edge if possible and, if not, 
was to be harried out of his beachhead by 
a counterattack on the night following the 
landing. "But," read the order, "in the 
eventuality we have been unable to expel 
the enemy . . . we will gradually fall back 
on our prepared positions in the southern 
part of the island and defend them to the 
last man." 44 

Whichever of the three possible beach 
areas was hit by the Americans, the bulk 
of the Japanese forces in the two other 
sectors was to rush to the point of attack 
and close arms with the invader. Tinian 
Town and Asiga Bay were strongly favored 
as the probable landing beaches, the north- 
west coast being relegated to third place 
in Ogata's list of priorities. 45 Thus, when 
the Americans chose this unlikely lane of 
approach, they achieved complete tactical 
surprise — a rare accomplishment in the 
Central Pacific, theater of war. 

43 4th Marine Div Representative Trans. 

43 NTLF Rpt Marianas, Phase III, Incl A, 
G-a Rpt. p. 23. 

44 Ibid., p. 9. 

45 4th Marine Div Representative Trans. 


Invasion and Capture 

Preliminary Bombardment 

Field Artillery 

The decision to land the assault infantry 
troops across beaches on northwestern Tin- 
ian had stemmed in part from a desire to 
make optimum use of artillery based on 
Saipan. In many Pacific landings such as 
those at Tarawa, Saipan, and Iwo Jima it 
had not been, or was not to be, practicable 
to seize neighboring islands for the purpose 
of establishing bases for field artillery 
before the principal landing operations. 
Hence reliance had to be put entirely on 
naval gunfire and aircraft for whatever 
preliminary bombardment was laid down. 
In these operations some part, and prob- 
ably a considerable part, of the casualties 
incurred by the infantry during the am- 
phibious assault phase must be attributed 
to the limitations inherent in naval and 
aerial bombardment. The invasion of Tin- 
ian, on the other hand, offered an ideal 
opportunity to supplement these arms with 
field artillery. Tinian was, in fact, favored 
with a more prolonged preliminary artillery 
bombardment than any other island jn the 
Central Pacific to be assaulted by Ameri- 
can troops. 

As early as 20 June, Battery B of the 
531st Field Artillery Battalion was ordered 
to emplace its 155-mm. guns to fire on 
Tinian, Lateral observation posts were 

established on southern Saipan and coun- 
tcrbattcry and destructive fires commenced 
forthwith. Four days later the battalion's 
other two batteries, relieved of their duties 
of supporting the troops on Saipan, 
turned around and began firing to the 
south. On 8 July, the day before Saipan 
was declared secured, the other three bat- 
talions of XXIV Corps Artillery were or- 
dered either to turn around in their present 
positions on southern Saipan or to displace 
to the area of Agingan Point and com- 
mence firing on Tinian. 1 Meanwhile, ob- 
servation planes made daily flights over 
Tinian to register fires and to accumulate 
intelligence data for future use. Observa- 
tion posts had been established on Agingan 
Point, Obiam Point, and Nafutan Point. 
All together, from 20 June through g July, 
while the troops were engaged in capturing 
Saipan, XXIV Corps Artillery fired a total 
of 331 missions — 7,571 rounds — on Tin- 
ian, or roughly a fifth of the total it ex- 
pended on Saipan. 2 

With Saipan officially secured on 9 July, 
even greater attention could be directed 
toward the island to the south. Corps ar- 
tillery increased the tempo of its bombard- 

1 The field artillery battalions were the 025th 
(155-mm. howitzers), the 145th (105-mm. how- 
itzers), and the 5|}2d ( 155-mm. guns). 

2 XXIV Corps Arty Final Rpt on Foraoer Opn, 
Phases I and III, S^2 Rpt, p. 8; Ibid., S-3 Rpt, 
PP. 7~ B - 



ment. By 15 July all four battalions of the 
27 th Division artillery had displaced to new 
positions to fire on Tinian, 3 as had also 
the two Marine divisions' howitzers (ex- 
cept for the two 75-mm. pack howitzer 
battalions of each) 4 and the 5th Battalion, 
V Amphibious Corps Artillery. This repre- 
sented a total of 156 pieces — two 155-mm. 
gun battalions, three 155-mm. howitzer 
battalions, and eight 105-mm. howitzer 
battalions/' Corps artillery had at its dis- 
posal nine organic planes plus the addi- 
tional observation planes of the units at- 
tached. Air activity became so heavy at 
Isely Field that a new airstrip for the ex- 
clusive use of observation planes had to be 
constructed somewhat to the westward. 

Except for a brief period on 16-17 July, 
when ammunition ran low, artillery main- 
tained a steady round-the-clock schedule of 
fire totaling 1,509 missions or 24,536 
rounds. In addition to counterbattery and 
harassing and general area bombardment, 
the artillery attempted to burn off cane 
fields with white phosphorus shells, but 
without much success because of the 
heavy rains that immediately preceded the 

For the most part, XXIV Corps Artillery 
confined its efforts to the area north of 
the line between Gurguan Point and Masa- 
log Point, while aircraft restricted their 
efforts to the southern half of the island. 
Naval ships were assigned any targets on 
Tinian deemed unsuitable to either of the 
other two arms. Co-ordination of the three 
supporting arms was assigned to the corps 
artillery representative attached to General 

s 27th Inf Div Arty Rpt Forager Opn, p. 25. 

1 Each Marine division had two 105-mm. how- 
itzer battalions and two 75-mm. pack howitzer 

5 XXIV Corps Arty Final Rpt on Forager Opn, 
Phases I and III, S-3 Rpt, p. 9. 

Schmidt's staff. In one instance, an artil- 
lery air observer discovered three 140-mm. 
coastal defense guns on Masalog Point 
that were within easy firing range of White 
Beaches 1 and q, but were masked from 
field artillery. The battleship Colorado was 
called in and, with its main batteries ad- 
justed by an airborne artillery observer, 
succeeded in neutralizing or destroying 
the enemy weapons. Because the spotting 
plane was not in direct radio contact with 
the ship, it was necessary for the plane to 
submit its spotting data to the artillery post 
by radio, whence they were relayed by tele- 
phone to General Schmidt's headquarters, 
and in turn by radio on another frequency 
to the firing ship. In spite of this some- 
what complicated system of communica- 
tions, the time lag was so slight as to be 
insignificant. 6 

Naval and Air Bombardment 

Naval guns had harrasscd Tinian inter- 
mittently since 1 1 June when Admiral 
Mitscher's Fast Carrier Force first made its 
appearance in the adjacent waters. For 
about a week after Saipan was secured, the 
job was left almost entirely to field artillery 
except for night fires delivered on Tinian 
Town by small naval craft. Then, starting 
on 15 July, naval gunfire was resumed. Ad- 
miral Hill's plan called for daily destructive 
fire by destroyers on specific targets that 
were not suitable to air or artillery, gradual 
intensification of daytime fire by employ- 
ment of additional destroyers and cruisers, 
and a continuation of night harassing mis- 
sions. Commencing on the 16th, seven de- 
stroyers began to deliver destructive fire 

p. 61. 

Ibid.; TF 52 Rpt Tinian, 24 Aug 44, Incl A, 



155-MM. Gun Firing at Tinian From Saipan during night bombardment of 
the island lo the south. 

against targets designated by corps artil- 
lery. On the evening of the 17 th, addi- 
tional destroyers commenced harassing fire 
against the Japanese who were known to 
be working feverishly at night to install 
beach defenses in the Asiga Bay area. On 
the 20th the cruiser Louisville was added 
to the armada, and two days later another 
cruiser, New Orleans, was sent into the 
line, as were two LCI gunboats whose job 
it was to pour 40-mm. fire into the many 
caves that pocked the cliffs along the shore 
line. 7 

Starting at 0600 on 23 July, the day be- 
fore the landing, Admiral Hill stepped up 

7 TF 52 Rpt Tinian, Incl A, pp. 58-59- 

his preparatory fire with a total of three 
old battleships, two heavy cruisers, three 
light cruisers, and sixteen destroyers, dis- 
tributed in such a way that the island of 
Tinian would be shelled from every point 
on the compass. The Navy made no effort 
to concentrate on White Beaches 1 and 2 
— in fact, the ships assigned to that area 
were fewer in number than in most of the 
other five sectors into which the waters off 
the island were subdivided. Deception was 
given even greater consideration than de- 
struction, and the naval gunfire plan can 
have given the Japanese no indicaton as 
to where the amphibious assault would 
take place. At 1 500 all ships ceased fire for 



an hour to permit aircraft to drop napalm 
bombs on the wooded area inland of Fai- 
bus San Hilo Point, and at 1720 naval fire 
was again discontinued for an air strike 
with napalm against the White Beaches 
area. During the night of 23-24 July de- 
stroyers and cruisers kept important road 
junctions between Faibus San Hilo and 
Gurguan Points interdicted, while the de- 
stroyer Norman Scott delivered harassing 
fire to cover Asiga Bay and all important 
road junctions on the cast side of the is- 
land. Another destroyer, Walter, covered 
a last-minute underwater reconnaissance of 
White Beaches 1 and 2 by Underwater 
Demolition Team 5, which unsuccessfully 
attempted to detonate some recently dis- 
covered mines. Throughout the day and 
thereafter during the operation, in order to 
minimize interference between air and 
naval gunfire, all ships were directed to 
deliver their fire from ranges of a maxi- 
mum ordinate of less than 1 ,000 feet when 
possible and to exceed that maximum only 
after notification by Admiral Hill. 8 

For air support on Tinian, Comdr. 
Lloyd B. Osborne, USN, who was Admiral 
Hill's Commander, Northern Support Air- 
craft, had 358 fighters, bombers, and tor- 
pedo planes under his control, mostly Army 
and Navy. Tinian, like Saipan, had of 
course felt the might of Marc Mitscher's 
carrier force early in June, and from then 
on was subject to an increasing tempo of 
bombing from naval planes of Turner's 
and Hill's escort carriers and from Army 
P-47's from Isely Field. Starting on 22 
June, P^47's of the 318th Fighter Group 
kept the airfields at Ushi and Gurguan 
Point and the new strip just east of Tinian 
Town under constant strafing and bomb- 

ing attacks. Tinian Town itself was reduced 
to rubble. 9 

On 22 July, just two days before the 
invasion, two P-47's dropped on Tinian 
the first napalm bombs used in the Pacific 
war. These were fire bombs consisting of 
jettisonable aircraft fuel tanks filled with 
a mixture of napalm gel and gasoline. 
Shortly before the scheduled landing on 
Tinian, Lt. Comdr. Louis W, Mang, 
USNR, recently arrived from the United 
States, easily persuaded Admiral Hill of the 
efficacy of these new bombs, and, since 
napalm was in short supply, an order for 
8,500 pounds was immediately dispatched 
to Pearl Harbor. Meanwhile, diescl oil was 
sometimes used as a less efficient substitute. 
The new fire bombs were found to be es- 
pecially effective in burning cane fields 
and underbrush. During the late afternoon 
of 23 July thirty were dropped immediately 
inland and on the flanks of White Beaches 
1 and 2 to burn off underbrush cover and 
destroy enemy personnel that might be lo- 
cated in open trenches and dugouts. In 
both respects the bombs were successful 
and their continued employment in the Pa- 
cific war was assured. 10 

The Landings 

At daybreak on 24 July a motley flotilla 
of ships and landing craft carrying the 2d 
and 4th Marine Divisions got under way 
from Tanapag Harbor for the short trip to 
Tinian, All together, it included 8 trans- 
ports, 37 LST's, 2 LSD's, 31 LCI's, 20 
LCT's, 92 LCM's, 100 LCVP's, 533 

' Ibid., pp. 62-63. 

9 This account of preliminary air activities is 
derived from TF 52 Rpt Tinian, Incl A, p. 26, 
and AAF Hist Div Army Air Forces in the Mari- 
anas Campaign, MS, OCMH. 

10 TF 52 Rpt Tinian, Incl A, pp. 93-96, 



LVT's, and 140 DUKW's. All of the 
LST's were assigned to the 4th Marine Di- 
vision, whose assault troops were nested in 
their amphibian tractors waiting for the 
moment when the bow doors would open 
and their vehicles would crawl out into the 
water and approach their line of depar- 
ture. Four of the LST's carried DUKW's, 
aboard which had been loaded four bat- 
talions of 75-mm. pack howitzers. Tanks 
were stowed in LCT's and LCM's, some 
to be carried to their destination in the 
wells of LSD's, others to get there under 
their own power. The eight transports car- 
ried the two regimental combat teams of 
the 2d Marine Division scheduled to make 
a diversionary feint at Tinian Town before 
landing in the rear of the assault troops. 
The division's other regiment would have 
to stand by on Saipan until ten of the 
LST's bound for White Beaches 1 and 2 
had unloaded and returned to pick it 
up. n \{Map76)\ 

On past the White Beaches to Sunharon 
Bay steamed the Demonstration Group — 
the transports carrying the two 2d Marine 
Division RCT's accompanied by their es- 
corts. 12 The Japanese expected a landing 
at Tinian Town, and Admiral Hill intended 
to prolong that expectation as far as pos- 
sible. Into the water went landing craft 
lowered from their mother transports; 
down the cargo nets climbed the marines 
of the 2d Division, to all appearances 
bound for the beach. 

Ashore, the Japanese reacted immedi- 
ately and furiously. Flashes from their 

guns, followed by tremendous geysers of 
water, kept the marines crouching low be- 
neath the gunwales of their boats as they 
approached the 2,000-yard line, which 
marked the inshore limits of their fake at- 
tack. Nothing but water and a few shell 
fragments hit the troops, but the escorting 
ships were not so lucky. The three British- 
made 6-inch coastal defense guns located 
south of Tinian Town struck the battleship 
Colorado and the destroyer Norman Scott, 
scoring twenty-two hits on the former and 
six on the latter. In this short action the 
Navy lost 62 killed and 245 wounded be- 
fore the shore battery could be silenced.' 3 
By 1 01 5 the boats and men had been 
recovered and the Demonstration Group 
stood up the coast toward White Beaches. 
The feint had been altogether successful. 
From Colonel Ogata to Tokyo went the 
message that more than a hundred landing 
barges had been repulsed in an attempt to 
get ashore at Tinian Town. 14 The 56th 
Naval Guard Force stuck to its guns that 
guarded Sunharon Bay, and no part of the 
3d Battalion, 50th Infantry Regiment, 
abandoned the southern sector to meet the 
amphibious landing in the north. 15 One 
Japanese infantryman probably reflected 
the thoughts and hopes of all when he 
wrote in his diary, "24 July: Today the 
enemy began to land on the beach at Tin- 
ian. 3 companies were sent out. Our pla- 
toon moved into position. . . . Up to 
0900 artillery fire was fierce in the direc- 
tion of Port Tinian but it became quiet 

11 Ibid., p. 6; NTLF Rpt Mananas, Phase III, 
InelH, LVT Rpt, pp. 1-3, 

12 Unless otherwise noted, the account of the 
diversionary feint against Tinian Town is derived 
from TF 52 Rpt Tinian, pp. 30-32; Hoffman, 
Tinian, pp. 43-45 ; Morison, New Cuinea and 
the Marianas, pp. 361-62. 

13 The battery was actually destroyed four days 
later by the battleship Tennessee. Morison, New 
Guinea and the Marianas, p. 36a. 

14 Japanese Studies in World War II, 55, pp. 

15 ad Marine Div Rpt Tinian, D-2 Rpt 76; 
NTLF Rpt Marianas, Phase I, Incl A, G-2 Rpt, 
p. 41. 

*M1 MO' Oit 

24 JUL* 



J 2 Mo- M 


g|2* \^ v^)8 



24 jLly-1 August 1944 


1000 ( 

3000 METERS 




after the enemy warships left. Maybe the 
enemy is retreating." 1C 

Meanwhile, the northwest coast of Tin- 
ian which had heretofore been treated with 
a studied impartiality, began to receive the 
full force of the attackers' armament. At 
0600 one battery of 155-mm. howitzers on 
Saipan commenced laying down a smoke 
screen on Mount Lasso to last for two and 
a half hours. Forty-five minutes later all of 
General Harper's guns and howitzers burst 
forth in a massed fire. For fifteen minutes 
artillery pounded known installations on 
northern Tinian, likely enemy assembly 
areas, and avenues of approaches to the 
beaches. Then, on signal, artillery lifted its 
fires to the woods and bluffs above the 
shore from which the Japanese might ob- 
serve the approach of the assault craft and 
vehicles. 17 

The chorus of destruction was swelled 
by Admiral Hill's support ships offshore. 
Starting at 0530, Tennessee, California, 
and Louisville opened up on White 
Beaches 1 and 2 with their main bat- 
teries, while Birmingham and Montpelier 
together bracketed Mount Lasso from op- 
posite sides of the island. Freshly arrived 
from Guam, Admiral Spruancc's flagship 
Indianapolis took over responsibility for 
Faibus San Hilo Point, which overlooked 
the landing beaches from the south. From 
0625 until 0640 all ships' fire ceased in or- 
der to allow an air strike against the 
beaches, and more particularly against the 
recently discovered mines off White Beach 
2, which neither underwater demolition 
swimmers, nor mine sweepers, nor ships' 

guns had been able to detonate. Thirty 
minutes before the scheduled landing time, 
artillery fire from Saipan shifted to the 
landing beaches. Throughout this period, 
one battery of 155-mm. howitzers fired a 
continuous barrage of smoke shells at 
Mount Lasso to prevent enemy observation' 
of the landing beaches. Then came the final 
crescendo of naval bombardment with de- 
stroyers abandoning their screening duties 
to add their five-inch shells to the general 

The first wave was to touch down at 
0730, but because of a slight delay in form- 
ing the landing waves, Admiral Hill or- 
dered a ten-minute postponement. The 
formation of landing craft and vehicles fol- 
lowed the pattern by now familiar in the 
Central Pacific. First in line abreast went 
the LCI gunboats, six toward White Beach 
1, nine toward White Beach 2. Astern 
came a wave of amphibian tanks, followed 
in turn by the troop-carrying amphibian 
tractors. Those bound for the northern 
landing beach were crewed by marines of 
the 2d Amphibian Tractor Battalion; the 
Army's 773d Amphibian Tractor Battalion 
carried the assault troops to the southern 
beach. As the gunboats approached shal- 
low water, they turned to port and star- 
board and took the flanks of the landing 
area under fire; at 300 yards the amphib- 
ian tanks turned toward the flanks and the 
first waves of amphibian tractors churned 
through the water to touch shore on White 
Beach 1 at 0742, on White Beach 2 at 
0750. 18 

By comparison with most assault land- 
ings in the Central Pacific, the initial in- 

10 4th Marine Div Representative Trans, Sec. 
VII, Diary of Takayoshi Yamazaki. 

17 XXIV Corps Arty Final Rpt On Forager 
Opn, Phases I and III, S-3 Rpt, p. 1 1. 

18 TF 52 Rpt Tinian, pp. 33, 65; XXIV Corps 
Arty Final Rpt on Forager Opn, Phases I and 
III, $-3 Rpt, p. 11. 



Invasion Craft make an easy landing on Tinian's While Beaches. 

vasion of Tinian was easy. 10 Within forty 
minutes after the touch down, the entire 
assault battalion of the 24th Marines was 
ashore on White Beach 1, and after a brief 
fire fight moved rapidly to the first objec- 
tive line. On White Beach 2, to the south, 
the 25th Marines faced somewhat heavier 
odds in the form of scattered land mines 
and two Japanese pillboxes undestroyed by 
the preliminary bombardment. Two LVT's 
were blown up by the mines. As the first 
wave of marines rushed inland it bypassed 
the pillboxes, which were subsequently re- 
duced without difriculty. Following the 
landing of the two assault regiments, tanks 
and half-tracks moved in on schedule, and 
in spite of the difficulties involved in land- 
ing over the narrow, cliff-flanked beaches, 

19 The account of JIG Day on Tinian is de- 
rived from the following sources; 4th Marine Div 
Opns Rpt Tinian, pp. [6-2O, 23-24; Hoffman, 
Tinian, pp. 48-58. 

none of the vehicles was lost. By midafter- 
noon DUKW's had succeeded in landing 
all four of the pack howitzer battalions as- 
signed to the 4th Marine Division, and by 
late afternoon the reserve regiment (23d 
Marines), after some delay caused by com- 
munications failure, was ashore and in po- 
sition in its assembly area on the right 
(south) flank of the beachhead. The two 
shore parties (1341st Engineer Battalion 
(Army) on White Beach 1 and the 2d Bat- 
talion, 20th Marines, on White Beach 2) 
had landed all their men and equipment 
by 1400. Two of the ten special portable 
EVT ramps were launched late in the 
afternoon. One capsized when the amphib- 
ian tractor struck a coral head, but the 
other was in proper position before night- 
fall. Six of the eight remaining ramps were 
installed the next day, the other two having 
been swamped in the process. 



4TH Marines Wading Toward White Beach i keep their weapons dry. 

By nightfall of the 24th the 4th Ma- 
rine Division had established a beach- 
head about 2,900 yards in width and 
almost a mile deep in the center. 20 Casual- 
ties for the day had been light — 15 killed 
and 225 wounded. 21 Barbed wire was 
strung along the entire length of the di- 
vision front, machine guns were emplaced 
to provide interlocking bands of fire, and 
pack howitzers were registered to cover the 
main road from Ushi Point airfield and 
other likely routes of enemy approach. 
Amphibian vehicles preloaded with ammu- 
nition had made their direct deliveries from 
ships lying offshore to the front-line troops 
as scheduled, and the latter were well sup- 
plied with reserve stocks of shells, mortars, 
and bullets. Every precaution was taken 
against the expected traditional first-night 

20 Hoffman, Tinian, Map 8, p. 3. 

- ' 4th Marine Div Opns Rpt Tinian, p. 24, 

enemy counterattack. When it came, the 
marines were ready for it. 

Japanese Counterattack 
24-25 July 

Having failed to stem the tide of the 
American assault over the beaches, Colonel 
Ogata now had to put into execution the 
second phase of his defensive plan — an 
organized counterattack during the first 
night after the landing. Whether the Jap- 
anese commander was in direct communi- 
cation with any of his troops other than 
those assigned to the Tinian Town area is 
uncertain, but his battalion commanders 
were well enough indoctrinated to launch 
the drive on their own initiative. Still under 
the illusion that the main amphibious as- 
sault would eventually be directed against 
Tinian Town, Ogata kept the 3d Battalion, 



50th Regiment, in position. The ad Bat- 
talion, to which had been assigned the 
northeast sector guarding Asiga Bay, also 
stayed out of the fight, as did the main 
body of the 56th Naval Guard Force 
manning the coastal defense guns along 
southern Tinian. Thus the brunt of the 
counterattack fell to Ogata's mobile reserve 
battalion (the 1st Battalion, 135th Infan- 
try Regiment), the 1st Battalion, 50th In- 
fantry, and sundry naval units stationed in 
the northern part of the island. 22 

The attack, when it came, consisted 
of three separate and seemingly un-co- 
ordinated thrusts against the American 
front — one along the western shore against 
the marines 1 left, one in the center at the 
boundary between the 25th and 24th Ma- 
rines, and a third against the 23d Marines 
on the American right flank. The first com- 
menced about 0200 and was undertaken 
exclusively by naval personnel coming 
down from the north, ft lasted almost five 
hours, but the Japanese failed to penetrate 
the marines lines at any point and lost an 
estimated 476 men in the effort, 23 

In the center, Japanese infantrymen 
struck about 0230. Between the two Ma- 
rine regiments they discovered a weak spot 
in the line through which a large body of 
Japanese poured and then branched out in 
two directions. One of the enemy groups 
turned west toward the rear areas of the 
25th Marines where it was eliminated after 
a brief fire fight. The other headed straight 
for the beach where it was eventually 

22 The account of the counterattack is derived 
from 4th Marine Div D-2 Periodic Rpt 72, and 
Hoffman, Tinian, pp. 62-68. 

' 23 This and subsequent figures for Japanese 
casualties are taken from dead counts made on 
the morning of 25 July. Since some of the Jap- 
anese may have been killed by preliminary naval, 
air, and artillery fire, the estimates are probably 

stopped by Marine artillerymen and ele- 
ments of the 2d Marine Division that had 
landed only a few hours earlier. The next 
morning almost 500 dead Japanese were 
counted in this area. 

The third attack was from the south and 
was preceded by five or six tanks, over half 
of the entire Japanese tank strength on the 
island. All the tanks were destroyed before 
they penetrated the lines of the 23d Ma- 
rines, against which the attack was di- 
rected. The infantrymen following the 
tanks had no better luck. In the eerie light 
furnished by naval star shells, the marines 
quickly disposed of this last group, esti- 
mated to number over 270 enemy soldiers. 

On the morning of the 25th a total of 
1,241 Japanese dead was counted, about 
700 identified as members of the two 
infantry regiments. Later interrogation of 
six prisoners of war revealed that by the 
morning of the 25th the 1st Battalion, 50th 
Infantry, had been virtually destroyed as a 
result of the fighting incident to the land- 
ing and the counterattack. 24 Another pris- 
oner of war testified that the ist Battalion, 
135th Infantry, Ogata's mobile reserve, 
had been "practically annihilated." a!i In 
the light of this evidence, General Gates' 
final conclusion seems irrefutable — in the 
early morning hours of 25 July, the 4th 
Marine Division "broke the Japs back in 
the battle for Tinian." 2B 

Capture of Northern Tinian 

During the next eight days, until the 
island was finally declared secured on 1 
August, the fighting on Tinian resolved it- 

' 2i 4th Marine Div Representative Trans, Note 
574, Sec. IX. 

2S 4th Marine Div D-2 Periodic Rpt 73, 26 
Jul 44. 

a * 4th Marine Div Opns Rpt Tinian, p. 25. 



self into three phases. First, it was neces- 
sary to push across to the eastern coast 
and seal off the entire northern third of 
the island, including such vital points as 
Mount Maga, Mount Lasso, and the Ushi 
Point airfield. Once this was accomplished, 
both Marine divisions could wheel to the 
south and proceed at a more rapid pace 
down Tinian's long axis until they reached 
the foot of the plateau that dominated the 
island's southern tip. Finally came the two- 
day battle for the plateau and the cliffs of 
Marpo Point that brought the operation to 
a close. 

On 25 July the 4th Marine Division, 
against only sporadic resistance, spent the 
day expanding the beachhead in all direc- 
tions. On the right, the 23d Marines 
covered about half the distance from White 
Beach 2 to Faibus San Hilo Point, meeting 
very few Japanese as they went. In the 
renter, the 25th and 24th Regiments made 
comparable advances in an easterly and 
southeasterly direction, the 25th captur- 
ing Mount Maga, which lay athwart the 
division's approaches to Mount Lasso, the 
highest point of the island. Army P-47's 
flying from Isely Field, as well as artillery 
based on Saipan, supported the action. On 
the left, the 8th Marines (attached to the 
4th Marine Division), assisted by tanks 
and by armored amphibians firing from the 
water, inched its way through the coral 
cliffs that lined the west coast north of 
White Beach 1. Meanwhile, most of the 
remainder of the 2d Marine Division had 
come ashore and by midafternoon General 
Watson had set up his command post in- 
land of White Beach 2. 27 

Marine casualties for the day were low, 
and enemy opposition, although occasion- 
ally fierce, was spotty. Nevertheless, those 
marines who did make contact with the 
enemy developed a healthy respect for the 
caliber of the Tinian garrison. The Japa- 
nese here were reported "to be better 
troops than those encountered on Saipan, 
with much better marksmanship." 28 

The fact is that immediately after the 
failure of his night counterattack, Colonel 
Ogata decided to disengage his forces and 
establish a new defense line running from 
Gurguan Point to the radio station inland 
from the center of Asiga Bay. The brief 
flurry of artillery fire that the 25th Marines 
had encountered during their approach to 
Mount Maga had merely been a delaying 
action. The bulk of the Japanese troops 
remaining in northern Tinian were with- 
drawing to the new line south of Mount 
Lasso. 20 

On JIG plus 2 (26 July) General 
Schmidt ordered the 4th Division, now on 
the right, to continue the attack in a 
southerly direction and the 2d Division to 
drive straight toward the east coast. 30 In 
the 4th Division zone, the 23d Marines 
pushed down the coast another 2,500 yards 
to a point well below Faibus San Hilo 
Point, while the 25th Marines occupied 
Mount Lasso, which had been entirely 
evacuated by the Japanese the day be- 
fore. 31 At Mount Lasso the 25th Marines 
reported evidence of "a careful, well- 
planned withdrawal, removing dead and 
destroying documents. Abandoned posi- 

27 This account is derived from 4th Marine 
Div Opns Rpt Tinian, pp. 25-26; 2d Marine 
Div Rpt Tinian, p. 2; Hoffman, Tinian, pp, 

28 4th Marine Div D-2 Rpt 72. See also 2d 
Marine Div D-2 Periodic. Rpt 74. 

is Japanese Studies in World War II, 55, p. 52. 

80 NTLF Opn Order 32-44. 

31 4th Marine Div Opns Rpt Tinian, pp. 26-27. 



Usui Point Airfield, in the northern portion of Tinian. This field was a 
major objective of the Tinian operation. 

tions had been well dug-in and carefully 
planned." 32 

On the corps left, meanwhile, the 8th 
Marines took over Ushi Point airfield, 
and its sister regiments sped on to the east 
coast and prepared to swing south. 33 
Thus, in three days, the major tactical ob- 
jectives of the Tinian invasion had been 
achieved: Ogata's major counterattack 
had been beaten off with a consequent loss 
to him of about one fourth of his force; 
Ushi Point airfield had been taken and was 
already in the process of being made oper- 
ational for American planes; Mount Lasso, 
the commanding position of the island had 
been occupied. In the words of the official 
Marine Corps historian, "Seldom was the 
victor of any of the Central Pacific con- 

32 4th Marine Div D-2 Rpt 73, Annex, D-2 
Material for Dispatch Summary 271700. 

33 ad Marine Div Rpt Tinian, pp. a-3. 

quests so unmistakably identified so early 
in the fight." 34 

Drive to the South 

From 27 through 30 July both Marine 
divisions made rapid advances toward the 
plateau that dominated the southern tip 
of the island. Enemy resistance on the 27th 
and 28th was almost nonexistent, but it 
gradually stiffened as the Americans ap- 
proached the Japanese last main defensive 
line. During the 27th and 28th General 
Schmidt employed what has been called an 
"elbowing" technique. That is, on the first 
day of the attack southward, he held back 
the 4th Division on the right while the 2d 
Division surged forward; the second day 
the roles were exchanged. The ostensible 

:M Hoffman, Tinian, p. 76. 



""fffc-5 7 . 


'" .T *>' 


Open Terrain of Ckntral Tinian, which permitted rapid advance toward 
the hill mass that dominates the southern tip of the island. 

purpose of this tactic was to permit his 
artillery to concentrate first in support of 
one division, then of the other. 35 

On the 29th this technique was aban- 
doned, and both divisions were ordered to 
advance as rapidly within their respective 
zones as conditions permitted. 36 On the 
right, the 4th Division on the 30th as- 
saulted a series of well-camouflaged cave 
positions on the west coast and after reduc- 
ing them pushed on in and through Tinian 
Town. Land mine fields on the town's out- 
skirts and along the beaches of Sunharon 
Bay slowed the advance a little, but the 
town itself had been reduced to rubble by 
naval gunfire and aerial bombardment and 
had been evacuated. 37 On the left, 

3S /tiU,p, 86. 

30 NTLF Opn Orders 35, 36. 
37 4th Marine Div D-a Periodic Rpt 77; Hoff- 
man, Tinian, pp. 96-97. 

the qd Division faced a tougher proposition 
as it came abreast the Masalog hill mass, 
but by nightfall of the 30th the area was 
overrun, and the Japanese were in retreat 
to the south. 38 

Apparently Colonel Ogata had relin- 
quished his mid-island line of defense al- 
most as soon as it had been established, 
and on the night of the 29th he moved his 
command post to a shrine in a cliff near 
Marpo Point. Seeing his delaying actions 
crumble before the advancing Americans, 
he ordered all Army and Navy forces to 
assemble on the southern tip of the island 
to defend the ridge line there. 39 

On their part, the Americans in four 
days had pushed their lines ahead about 

3S 21I Marine Div Rpt Tinian, pp. 3-4; Hoff- 
man, Tinian, pp. 98-100. 

39 4th Marine Div Representative Trans, Sec. 
IX, p. 5 . 



155-MM. Howitzer Emplacement on Tinian 

10,000 yards on the left and 4,000 on the 
right. 40 Coming so soon after the prolonged 
deadlocks of the Saipan battle, this seemed 
indeed like a sprint. As General Cates ex- 
pressed it, the marines were "heading for 
the barn." 41 

One of the reasons for this rapid move- 
ment was the gently undulating terrain of 
central Tinian, which permitted tanks to 
be used with far greater effectiveness than 
had been the case on Saipan. To each reg- 
iment was assigned one reinforced medium 
tank company (eighteen tanks), a platoon 
of four flame thrower tanks, and two light 
tanks. Each of the tank companies stayed 
with its parent regiment throughout the 
operation, which of course facilitated tank- 
infantry co-ordination. Also, one infantry 
regiment of each division was at all times 
in reserve, thus giving its assigned tank 
unit an opportunity to repair its ve- 
hicles. In addition, communications be- 

tween tanks and infantry were markedly 
improved over those on Saipan, 4a 

Tinian's terrain also offered more favor- 
able opportunities for the employment of 
artillery than had Saipan's. Initially, all ar- 
tillery in support of the 4th Marine Di- 
vision was based on Saipan except for four 
battalions of 75-mm. pack howitzers from 
the 14th and 10th Marine Artillery Regi- 
ments, On 26 July the 3d Battalion, 
14th Marines (105-mm. howitzers), came 
ashore, followed the next day by the 105- 
mm. howitzers of the 4th Artillery Battal- 
ion, V Amphibious Corps (attached to the 
4th Marine Division), and the two 105- 
mm. howitzer battalions of the 10th Ma- 
rines. 43 On 28 July the 419th Field 
Artillery Group of the XXIV Corps Artil- 
lery (155-mm. howitzers) displaced from 

40 See Hoffman, Tinian, Map 10, p. 87. 

41 4th Marine Div Opns Rpt Tinian, p. 29. 

'- Ibid., Annex G, Opns Rpt, p. 19; Ibid., An- 
nex K, 4th Tank Bn Rpt. 

13 4th Marine Div Opns Rpt Tinian, Annex F, 
14th Marines Rpt, pp. 4-6; 2d Marine Div SAR, 
Phase III, Foragkr, Annex, 10th Marines Rpt, 
PP- i-3- 



Saipan to positions on Tinian, as did one 
battery of the 106th Field Artillery Bat- 
talion (also 155-mm. howitzers). The 
other two batteries of this battalion were 
forced to return to their Saipan positions 
since they were unable to land over the 
White Beaches because of wrecked ponton 
causeways there. The XXIV Corps Ar- 
tillery's 420th Field Artillery Group re- 
mained on Saipan throughoat, its 155-mm. 
guns having sufficient range to hit any part 
of the southern island. 44 

Corps artillery alone fired j ,404 missions, 
totaling 46,309 rounds, during the assault 
and capture of Tinian. 45 Added to this, of 
course, was the Marine divisions' or- 
ganic artillery, which fired approximately 
142,000 additional rounds. 48 As one Jap- 
anese prisoner testified, "You couldn't 
drop a stick without bringing down ar- 
tillery." 47 

Close air support on Tinian was pro- 
vided by Army P-47's flying from Isely 
Field, as well as Navy carrier-based planes. 
As the two Marine divisions started their 
drive south from the Mount Lasso line, 
Army and Navy aircraft alternated in pro- 
viding air cover in advance of the troops, 
averaging 175 sorties daily. Admiral Hill's 
support aircraft commander, Commander 
Osborne, had the final responsibility for 
approving and directing all air strikes re- 
quested by the Marines on the front line. 
Under him was the Commander, Support 
Aircraft Ashore, who was stationed on Isely 
Field with authority to exercise direct con- 
trol over the P-47's. Final responsibility 
for co-ordinating air, artillery, and naval 

41 XXIV Corps Arty Final Rpt on Forager 
Opn, Phases I and III, S-3 Rpt, p. 12. 

45 Ibid. 

4H NTLF Rpt Marianas, Phase III, In<.l F, 
G-4 Rpt, and Incl D. 

47 4th Marine Div Representative Trans. 

bombardment resided with a representative 
from XXIV Corps Artillery at General 
Schmidt's headquarters. 48 

Only in rare instances were supporting 
aircraft directly coached into their targets 
by air liaison parties attached to the in- 
fantry. The Navy command was fearful of 
turning over control of supporting aircraft 
to the ground troops for a variety of rea- 
sons. Neither the air liaison parties nor the 
pilots were deemed sufficiently trained in 
the niceties of air-ground co-ordination to 
risk it. Air-ground communications were 
uncertain chiefly because of unsatisfactory 
and insufficient radio equipment. Finally, 
the danger of decentralizing control of air 
strikes over such a small target as Tinian 
was considerable. Once the drive to the 
south was under way, the marines had on 
the front lines at all times at least twelve 
battalions, each with its own air liaison 
party. The lines themselves were often ir- 
regular and of course the front narrowed 
as the troops approached the southern tip 
of the island. Under these conditions, to 
have allowed each battalion to control its 
own called strikes would have seriously en- 
dangered the units on the flanks, and the 
risk was considered unacceptable. 49 

On Tinian, as on Saipan, the time lag 
between requests for and execution of air 
strikes was a cause for dissatisfaction 
among the ground troop commanders. 
Even when the planes were on station 
above the target, half an hour was usually 
required to complete an air strike, and 
when the planes had to be flown from their 
mother carriers or from Isely Field, an 
hour's delay was more common. One air 
liaison party had to wait a full nineteen 

4S TF 56 Rpt Forager Tinian, Incl I; NTLF 
Opn Plan 3-44, Annexes C, D, F. 
49 Hoffman, Tinian, p. 128. 



hours to get its request honored, but that 
was exceptional. 00 Nevertheless, General 
Cates and General Watson were both of 
the opinion that the execution of air strikes 
was several cuts above what it had been on 
Saipan — largely because the pilots had 
been better briefed, were more familiar 
with the terrain, and were gaining ex- 
perience. 51 

Naval fire support during the battle for 
Tinian was also considered to be an im- 
provement over that for Saipan, again be- 
cause of greater experience on the part of 
ships' companies and because of more 
favorable terrain. Preparation fires were 
commonly delivered on request of division 
commanders before the morning jump-off 
to supplement the field artillery, and 
counterbattery interdiction and destructive 
fires were delivered daily on call. 52 In the 
opinion of Lt. Col. E. G. Van Orman, 
USMC, who was Holland Smith's naval 
gunfire officer, "In the occupation of Tin- 
ian call fire procedure was carried out 
much more satisfactorily than at either 
Saipan or Guam because of experience 
gained by all hands at Saipan and ex- 
changed and clarified in meetings of all 
personnel both afloat and ashore prior to 
J-Day." 53 

Perhaps the most unusual features of the 
battle for Tinian were the techniques that 
were improvised for getting supplies over 
the narrow beaches to the front-line troops 
both during the initial amphibious phase 
and later. Responsibility for preparing the 
beaches themselves and for controlling traf- 

fic over them fell to the Army's 1341st 
Engineer Battalion and the 2d Battalion, 
20th Marines (the Engineer regiment of 
the 4th Marine Division ) . The former was 
assigned to White Beach 1, the latter to 
White Beach 2. Both were landed on JIG 
Day, at first operating under control of the 
4th Marine Division and later (26 July) 
coming under the direct command of Gen- 
eral Schmidt's shore party officer, Col. 
Cyril W. Martyr, USMC. 54 Also on the 
a6th the 2d Battalion, 18th Marines (or- 
ganic to the 2d Marine Division), came 
ashore to assist at White Beach q, to work 
in the division dumps, and later to help 
unload aircraft at Ushi Point airfield.* 5 

To facilitate unloading on the White 
Beaches, two ponton causeways were as- 
sembled on Saipan and on the afternoon 
of 24 July were towed to Tinian. There 
they were put to excellent use until the 
night of 29 July, when the tail of a typhoon 
that had been building up in the Philippine 
Sea hit Tinian with full force. The storm 
broached one of the artificial piers and 
broke the other in two/' 6 

It was during this typhoon that the 
DUKW's once again demonstrated their 
outstanding versatility and durability, A- 
bout half of the 140 amphibian trucks 
used on Tinian were crewed by Army per- 
sonnel of the 477th Amphibian Truck 
Company and the 27th Division Provisional 
Amphibian Truck Company, the rest were 
crewed by marines of the 1st and 2d Ma- 
rine Amphibian Truck Companies. As the 
seas mounted on the afternoon of 29 July, 

50 4th Marine Div Opns Rpt Tinian, Annex 
E, Special Comments and Recommendations. 

st ad Marine Div Rpt Tinian, p. 18; 4th Ma- 
rine Div Opns Rpt Tinian, Annex C, p. 1 2. 

03 TF 56 Rpt Forager Tinian, Incl I, Naval 
Gunfire Rpt. 

™ Ibid., p. 138. 

54 4th Marine Div Opns Rpt Tinian, Annex G, 
20th Marines Rpt, pp. 1-4; NTLF Rpt Marianas, 
Phase III, Annex J (2), Engineer and Shore 
Party Rpt, p. 1. 

SH 2d Bn 1 8th Marines Rpt, p. 1. 

SB TF 56 Rpt Forager Tinian, Incl J, Engi- 
neer and Shore Party Rpt, p. 2, 



broaching one LST and washing a landing 
craft control boat up on the beach, it de- 
veloped that of all the small craft and ve- 
hicles present, only the DUKW's were sea- 
worthy enough to operate in the heavy 
swells, and for the duration of the storm 
they were solely responsible for overwater 
supply/' 7 

As General Schmidt's supply officer re- 
marked, the DUKW's at Tinian "per- 
formed an astounding feat of supply." 58 
Equipped with A-frames, they carried most 
of the artillery pieces from LST's directly 
to firing positions ashore. 159 They were 
solely responsible for averting a serious fuel 
shortage when the typhoon struck, since 
they were the only means at hand for get- 
ting through the surf to fuel barges an- 
chored off the northwest coast of Tinian. 
Not only did they prove more seaworthy 
than their sister amphibian vehicle, the 
LVT, but over Tinian's fairly well 
developed road system they delivered sup- 
plies more quickly to the troops as they 
approached the southern end of the island 
and also, of course, wrought much less 
damage to the roads than did the tracked 
vehicles, 00 

The DUKW's and LVT's bore the main 
burden of shuttling supplies and equipment 
of all kinds from the vessels lying off the 
beaches straight to inland dumps, or even 
to the men on the front lines as they 
pushed farther southward. It was this di- 

57 NTTF Rpt Marianas, Phase III, Incl F, 
G-4 Rpt, Sec. B, pp. 2-3. To supplement this 
delivery, transport planes from Saipan flew ra- 
tions to the front-line troops on Tinian. 

D *NTLF Rpt Marianas, Phase III, G-4 Rpt, 
Sec, B, pp. 2-3. 

r ' 9 This particular use of the specially adapt- 
ed DUKW's had been introduced by the 7th In- 
fantry Division at the invasion of Kwajalein. Sec 
Growl and Love, Gilberts and \larshalls, p. 227. 

(i " Hoffman, Tinian, p. 136, 

rect and rapid system of supply, which 
eliminated manhandling supplies on the 
beaches, that struck most observers as the 
outstanding feature of the Tinian battle. 
In the words of General Schmidt's supply 
officer : 

This operation was in many ways a re- 
markable demonstration of the fact that pre- 
conceived notions and amphibious doctrine 
can be altered radically on the spot. In 
effect, a reinforced corps was landed over 
less than aoo yards of beach and over a diffi- 
cult reef, and was supplied throughout nine 
days of heavy combat without handling so 
much as one pound of supplies in the usual 
shore party manner. Everything rolled in on 
wheels. When a violent sea made impossible 
the landing of trucks, the DUKW's took over 
all supply, supplemented to a minor degree 
by incoming air evacuation planes bringing 
in rations. The troops never lacked what they 
required at the time it was required. 61 

Tinian Secured 

Nightfall of 30 July found the two Ma- 
rine divisions drawn up on a line just north 
of the hill mass that dominated the south- 
ern tip of Tinian. South of them was about 
a mile of flat land that terminated in an 
abrupt wooded escarpment rising to a pla- 
teau. Here Colonel Ogata had elected to 
make his last desperate stand. The area 
consisted mostly of an oblong mountain 
mass about 5,000 yards long and 2,000 
wide running generally in a northeast- 
southwest direction. This high ground was 
something like a huge mesa with the steep 
ridges and cliffs of its shoulders supporting 
the comparatively gentle slopes along the 

(il NTLF Rpt Marianas, Phase III, Incl F, 
Supply Rpt, p. 4. This statement contains two 
minor inaccuracies. The beaches were somewhat 
more than aoo yards in width, and the vehicles 
that carried supplies inland were tracked as well 
as wheeled. 



top. Of the two long sides, one faced the 
flat land around Tinian Town, the other 
met the sea on the east coast. The entire 
southern tip sloped steeply to the water. 
Colonel Ogata's defense line was drawn on 
the forward (northwestern) slopes of the 
hill mass. 

On the morning of 31 July the marines 
attacked as before with two divisions 
abreast, the 4th on the right, 2d on the 
left. Before the jump-off, two battleships 
(Tennessee and California) and three 
cruisers (Louisville, Montpelier, and Bir- 
mingham) fired about 615 tons of shells 
into the area, and Army bombers dropped 
about 69 tons of explosives. 62 

On the right the 24th Marines, sup- 
ported by tanks and armored amphibians, 
made slow but steady progress against 
stiffening resistance along the coast line- 
south of Tinian Town, and by the end of 
the day had advanced about 2,500 
yards. 63 The 23d Marines on the division 
left faced greater obstacles as it came up 
against the cliff line that marked the north- 
western face of the plateau. With the help 
of supporting tanks, the regiment knocked 
out a 47-mm. antitank gun in the path of 
its progress and at day's end dug in at the 
foot of the cliff, though one company 
reached the top and spent the night 
there. 64 

On the left, the 2d Marine Division at- 
tacked with three regiments abreast. The 
two left regiments, 2d and 6th Marines, 
moved forward to the base of the cliff 
against only light rifle and machine gun 
fire, but the 8th Marines was not let off 
so easy. In its zone lay the precipitous 
double-hairpin road that offered the most 

B2 TF 52 Rpt Tinian, pp. 79-80, 132. 
60 Hoffman, Tinian, Map I r, p. 103. 
fil 4th Marine Div Opns Rpt Tinian, pp. 30-31 ; 
Hoffman, Tinian, pp, ioa-07. 

feasible route to the top of the plateau, 
and to get even partial command of this 
artery took a day of heavy fighting and 
arduous climbing. By late afternoon one 
company had reached the top of the cliff, 
followed after dark by most of the two as- 
sault battalions of the 8th Marines, but, as 
night fell, there was a gap of 600 yards on 
the right and one of 350 yards on the left 
of the Marine battalions atop the cliff. The 
time and the situation were ripe for a Jap- 
anese counterattack, and it came as ex- 
pected.' 5 '' 

Colonel Ogata personally led the 
counterattack, which was directed mainly 
against the 8th Marines atop the cliff. Ac- 
cording to one Japanese prisoner of war, 
Ogata was killed during the charge by 
American machine guns and was last seen 
hanging dead over the Marines' barbed 
wire.' 56 About 2300 the Japanese first 
struck elements of the 8th Marines but 
were repulsed. Three hours later a force 
of some 150 of the enemy suddenly rushed 
the hairpin road up which the marines had 
been trying to carry ammunition and other 
supplies. There, the Japanese set up a road- 
block, burned two ambulance jeeps, and 
threatened to cut off the two American 
battalions on top of the plateau. An hour 
later a platoon of the enemy moved up the 
road and attacked from the rear. In a short 
but furious fire fight they were driven back 
and their roadblock was eliminated. For 
the next three hours, 75-mm. half-tracks, 
mortars, and field artillery kept the enemy 
at bay, but at 0515 came the final banzai 

b: > %d Marine Div Rpt Tinian, pp. 5-fi ; Hoff- 
man, Tinian, pp. 107-10. 

r ' fi NTLF Rpt Marianas, Phase III, Incl D, 
G-2 Rpt, pp. 1 8, 20, 21. A Japanese study of 
the operation prepared after the war, on the other 
hand, has Ogata still alive and leading another 
counterattack as late as 3 August. Japanese 
Studies in World War II, No. 55, p. 53. 



charge. For a full half hour the attackers 
charged the Marine lines, but at no point 
did they penetrate. Daybreak revealed over 
a hundred enemy dead in an area only 
about 70 yards square. Later interrogation 
of prisoners indicated that the entire coun- 
terattacking force had numbered between 
six and eight hundred. 67 

With the failure of this last counterat- 
tack, organized Japanese resistance quickly 
came to an end. By late afternoon on 1 
August both Marine divisions had reached 
the southern edge of the cliff line, and at 
1855 General Schmidt declared the island 
secured.' 18 Mopping up, to be sure, was a 
long and often bloody process. 69 Not until 
1 January 1945 were the remnants of the 
enemy force considered sufficiently dis- 
posed of to permit the mop-up troops, the 
8th Marines, to be transferred to Saipan. 
In the three months after Tinian was 
turned over to the island commander, Maj. 
Gen. James L. Underhill, USMG, a total 
of 542 Japanese were reported killed. 70 

All together, the capture of Tinian had 
cost the invading ground forces a total of 
328 killed and 1,571 wounded in action, 
almost all of them Marine Corps person- 
nel. 71 In exchange, the Japanese sacrificed 
their entire garrison of more than eight 
thousand men, most of them killed. More 
significant than this death toll was the fact 
that the U.S. forces had succeeded in 

" 7 Hoffman, Tinian, pp. 109-13. 

HS NTLF Rpt Marianas, Phase III, p. 15. 

69 Pfc, Robert L. Wilson of the ad Battalion, 
6th Marines, was posthumously awarded the 
Medal of Honor for covering with his own body 
a live hand grenade on 3 August, two days after 
Tinian had been declared secured. For a similar 
feat performed on 30 July, Pvt. Joseph W. Oz- 
bourn, 1st Battalion, 23d Marines, was given the 
same award. Hoffman, Tiniarij pp. 98, 1 j 7, 

70 Hoffman, Tinian, p. 121. 

71 Ibid., App. Ill, p. 150. 

wresting from the enemy one of the best 
airfield sites in the Central Pacific. Ushi 
Point airfield and Gurguan Point airfield, 
enlarged and expanded, became vital bases 
for the XXI Bomber Command, which in 
the spring and summer of 1945 would un- 
leash its very long range bombers against 
the Japanese homeland with such devasta- 
ting effect. Significantly, it was from Ushi 
Point airfield that the B-qq Enola Gay, 
took off on 6 August 1945 to drop the first 
atomic bomb on the city of Hiroshima. 72 
Tinian was largely a Marine Corps show. 
A Marine headquarters made the tactical 
plans and Marine infantrymen carried the 
main burden of attacking and overrunning 
the island. Nevertheless, the Army's role 
was by no means negligible. In the ship-to- 
shore movement, over half of the amphib- 
ian tractors were provided by the Army 
and crewed by soldiers. Half of the amphib- 
ian trucks that landed the artillery and 
later, during the typhoon, played such an 
important role in supplying the troops, 
were Army-manned. An Army engineer 
battalion acted as shore party for one of the 
landing beaches. Army artillery played a 
decisive part in the preliminary bombard- 
ment and in supporting the marines after 
they had landed. Finally, Army P-47's flew 
continuous close support missions in front 
of the Marine infantry. Since so much of 
the success of the operation depended upon 
artillery based on Saipan, on the efficient 
work of the amphibian vehicles, on the or- 
ganization of the supply system at the 
beaches, and on close air support, h can 
be concluded that the Army's share in the 
reduction of Tinian was far out of propor- 
tion to the number of its personnel actually 
committed to the operation. 

72 Craven and Cate, AAF V, pp. 715-17. 



Plans and Preparations 

The Island 

To the invaders of Guam, southernmost 
of the Marianas chain, the physiography of 
the island presented essentially the same 
problems and challenges that had already 
been encountered at Saipan — those for 
Guam were just on a larger scale. Located 
a little more than a hundred miles south of 
Saipan, Guam is more than twice its size, 
measuring 228 square miles in area. From 
Ritidian Point in the north to the southern 
coast line, the distance is about thirty-four 
miles; the width o f the island varies from 
five to nine miles, 1 

(Map V.\ 

Guam, like Saipan, is surrounded by 
coral reefs ranging in width from 25 to 700 
yards. Even the lowest of these is covered 
at high tide by only about two feet of water 
— a condition that of course made the em- 
ployment of amphibian tractors mandatory 
in the projected ship-to-shore movement. 
Around the entire northern half of the is- 
land from Fadian Point on the east coast 
to Tumon Bay on the west, sheer cliffs ris- 
ing to 600 feet ruled that area out for 
landing. In the southern part of the island 
the shore line cliffs are somewhat less for- 
bidding, but even so in many places, such 
as at the tip of Orote Peninsula on the 

1 This account of the physical features of Guam 
is derived from ONI 99, and Military Intelligence 
Service, War Department (MIS WD), Survey of 
Guam, 1943. 

west coast, they are still too precipitous to 
permit rapid movement inland by any large 
numbers of men approaching from the sea. 
The southern and southeast coasts, exposed 
as they are to the prevailing easterly winds, 
are pounded too heavily by surf to permit 
easy landing operations. This leaves about 
fifteen miles of coast line feasible for an 
amphibious assault, all on the west coast, 
north and south of Orote Peninsula. At 
various places in this region, the reef is 
low enough and the sandy beaches are both 
wide and deep enough to permit invading 
troops to get ashore and establish a foothold 
before assaulting the mountainous terrain 

Although nowhere does Guam's moun- 
tain range reach the heights of Mount Ta- 
potchau on Saipan, it still presents obsta- 
cles of no mean proportions. The northern 
part consists almost entirely of a coral lime- 
stone plateau broken by three elevations, 
Mount Barrigada (674 feet), Mount Santa 
Rosa (870 feet), and Mount Machanao 
{610 feet ) . The central part, the waist of 
the island between Agana Bay and Pago 
Bay, is mostly lowland draining into the 
Agana River through a wide swamp of the 
same name. Just south of the waist the land 
begins to rise again toward the mountain 
range that runs to the southern tip of the 
island. Dominating the northern part of this 
range are Mount Ghachao, Mount Alutom, 



Cliff Link at Tip of Orote Peninsula 

and Mount Tenjo, all inland from Apra 
Harbor and all attaining more than a 
thousand feet. East of Agat Bay below 
Orote Peninsula lies Mount Alifan (869 
feet) ; south of it and inland from Facpi 
Point is Mount Lamlam, the highest point 
on the island ( 1,334 feet). 2 

Though Guam's mountain mass is not so 
high as Saipan's, its vegetation is lusher, 
heavier, and thicker. A degree and a half of 
latitude in this area of the world makes a 
difference, and Guam is considerably more 
tropical than the northern island. At the 
time of the invasion the northern section 
of Guam was heavily covered with tropical 
forests, weeds, trailing vines, lianas, air 
plants, and underbrush — all combining to 
make foot passage almost impossible ex- 
cept through man-made jungle trails. The 
mountain tops themselves were mostly bar- 

s Heights of land are, derived from Army Map 
Service, Map, Island of Guam, Scale 1:62,500, 
Washington, 1944, 

ren volcanic rock covered only with sparse 
growths of sword grass and scrub. The 
southern plateau was covered mostly with 
sword, cogon, and bunch grass and scrub 
forest, except between Mount Alifan and 
Mount Lamlam, where timber grew in 
fairly large stands. 

To facilitate passage over and through 
this rough and forbidding country there 
were, in the summer of 1 944, about a hun- 
dred miles of hard-surface road, linked to- 
gether by single-lane unsurfaced roads and 
a network of narrow jungle trails cut 
through the bush. The main road ran from 
the town of Agat along the west coast to 
Agana, then northeast to Finegayan, where 
it split into two parallel branches, each 
terminating near Mount Machanao near 
the northern tip of the island. Another 
branch of the same road ran northeast to 
the village of Yigo, where it dwindled into 
a narrow unsurfaced road that continued 
on almost to Pati Point, on the northeast- 


AP~H„,,„, S Um0 » 


Orote Peninsula 

ern coast. Also from Agana to Pago Bay on 
the east coast stretched a main artery that 
continued south and west along the coast 
line to Umatac, Umatac and Agat on the 
west coast were connected only by a dirt 

Except for the surfaced highways, the 
roads and trails were normally all but im- 
passable during the rainy season, which 
lasted from July to November. During this 
summer monsoon period, 20 to 25 days out 
of each month were rainy. Mean tempera- 
ture was about 87 ° Fahrenheit and aver- 
age humidity about 90 percent— factors 
that would increase the discomfort of com- 
bat troops, whether American or Japanese. 

Plans for the Invasion 

Guam was initially included in the list 
of American targets for 1944 by the Joint 
Chiefs of Staff directive of 12 March 1944 
that ordered Admiral Nimitz to prepare to 
seize and occupy the southern Marianas. 

Like the islands to the north, it offered 
sites for B-29 bases and, in addition, Apra 
Harbor was the best ship anchorage in the 
entire archipelago, having excellent possi- 
bilities for development into a small for- 
ward naval base. Then too, Guam, like the 
Philippines, had been an American posses- 
sion; its native population was presumed 
loyal to the United States, and its libera- 
tion deemed a moral obligation. 

Little more than a week had passed since 
the 12 March directive when Admiral 
Nimitz issued a preliminary order (dated 
20 March) for the seizure of the southern 
Marianas, including Guam. Saipan and 
Tinian were assigned to the V Amphibious 
Corps. To the III Amphibious Corps, 
commanded by General Geiger, USMC, 
went the job of recapturing Guam. 3 Gen- 

3 Actually at this date the unit was named I 
Marine Amphibious Corps (I MAC), but on 15 
April its designation was changed to III Amphi- 
bious Corps. To avoid confusion the latter title 
is used throughout this volume. 

Chart 2 — Task Organization for Major Commands for Attack on Guam 

Task Fo*ce 50 
Fifrh Fleet 

Adm. R. A. Spruance 




Tailc Forte 58 
FqsI Carrier Force 

Vie* Adm, M. A. Mttscher 

Task Fore* 51 

Joint Expeditionary Foiee 

Vice Adm, R. K. Turner. 

Task Force 57 

Forward Area, Cenlrol Pacific 
(Land-Based Aircraft) 
Vice Adm, J. H, Hoover 

Task Farce 56 

Expeditionary Troop* 

Li Gen.H. M. Smith, USmC 

-i — +-■ 

Task Force 53 
Southern Attack Fatce 
Rear Adm. R. L. Conolly 

Task Group 56,2 

Southern Troopi ond Landing Farce 

Moj. Gen. H. S. Geiger, U$MC 


III Amphibious Corps Troop* 
Maj. Gen. R S. Geiger, USMC 

IN Amphibious Corps Artillery 
Brig, Gen, P, A. del Voile, USMC 

3d Marine Division 

Moj* Gen. A. H. Turnoge. USMC 

I ft Proviiional Marine Brigade 

Brigj. Gen. L. C. Shepherd, Jr., USMC 

77th Infantry Division 

Moj. Gen, A. D. Bruce, USA 

To*k Group 53.1 

Northern Attack Group 
Rear Adm. R. L. Conolly 

Task Grouo 53.3 
Northern Transport Group 

Copt. P, Buchanan 

Task Group 53,16 
Tractor Group Three 
Capt, G, B, Carter 

Task Group 53.9 

Minesweeping and HydrOgraphic Group 

Lt. Comdr. G- M. Esfep 

Task Group 53.2 
Southern Attack Group 
Rear Adm. L. F. Reiftnider 

Task Group 53.4 
Southern Transport Group 
Capt + J. B. McGavern 

Task Group 53.1 7 
Tractor Group Four 

Comdr. E. A. McfolJ 

Task Group 53.6 

Minesweeping and Hydrographie Unit 

Comdr. R. R. Sampson 

Task Group 53.5 

Southern Fire Support Group 

Rear Adm. W. L. Amsvrorth 

Task Group 53.7 
Carrier Support Group 
Rear Adm + V- H. Rogsdale 

Task Group 53.19 
Corpt Reserve Group 
Capt. H. B. Knowles 

Command and Opemhonal conltol 

Opcic- Onal conlral only until troops are 
citabluhrd ashore 



eral Geiger was to have under his com- 
mand the 3d Marine Division; the 1st Pro- 
visional Marine Brigade, consisting of the 
4th and 22d Marine Regiments, reinforced; 
III Amphibious Corps Artillery; and the 
9th and 14th Marine Defense Battalions. 
The 27 th Infantry Division was constituted 
Expeditionary Troops Reserve for the en- 
tire force of the two corps. The 77th In- 
fantry Division, commanded by Maj. Gen. 
Andrew D. Bruce — still in the United 
States but scheduled shortly to move to 
Hawaii — was designated Area Reserve. 
Twenty days after Saipan was assaulted, 
the 77 th was alerted for movement into the 
Marianas, 4 

Command relationships between the top 
commanders for the Guam phase (Phase 
III) of the Marianas operation were to be 
in every way similar to those that were ob- 
tained for Saipan and Tinian. Under Ad- 
miral Nimitz, Admiral Spruance as Com- 
mander, Fifth Fleet, was in over-all com- 
mand. Under him came Admiral Turner, 
Commander, Joint Expeditionary Force 
(Task Force 51), and General Holland 
Smith, Commander, Expeditionary Troops 
(Task Force 56), whose respective powers 
and responsibilities on this echelon of com- 
mand have already been described/' The 
Joint Expeditionary Force was in turn di- 
vided into two groups. The first, called 
Northern Attack Force (Task Force 52), 
also under Admiral Turner, was directed 
to land and support the assault troops on 
Saipan and Tinian. The second, designated 
Southern Attack Force, commanded by 
Admiral Conolly, USN, was given the same 
task for Guam. In like manner, General 
Holland Smith's Expeditionary Troops was 

Maj, Gen. Roy S. Geiger 

split into two parts: Northern Troops and 
Landing Force (Task Group 56.1 ) consist- 
ing mainly of the V Amphibious Corps plus 
the XXIV Corps Artillery and commanded 
also by General Smith was allocated to Sai- 
pan and Tinian; Southern Troops and 
Landing Force (Task Group 56.2), made 
up mostly of the III Amphibious Corps, 
commanded by General Geiger, was des- 
tined for Guam. The command relation- 
ships between General Geiger and Admiral 
Conolly were essentially the same as those 
that obtained between General Smith as 
Commander, Northern Troops and Land- 
ing Force, and Admiral Turner as Com- 
mander, Northern Attack Force. Thus, 
during the ship-to-shore movement, Con- 


TF 56 Rpt Forager, IncI B, G-5 Rpt, pp. 
See above, p. 33. 

" TF 56 Opn Plan 2-44; 11 Apr 44; Opn Plan 
y-44, 26 Apr 44; III Phib Corps Opn Plan 1-44, 
I I May 44, Annex King. 



oily was to command the landing force 
through Geiger. Once Geiger determined 
that the status of the landing operation 
permitted;, he was to assume command of 
the troops on shore and r eport that f act to 
the task force commander .1 ( Chart 2 ) I 

Planning for the invasion of Guam was 
somewhat complicated by the vast dis- 
tances that lay between the headquarters 
of the various commanders concerned. 
General Gcigcr's III Amphibious Corps 
headquarters was located at Guadalcanal; 
General Holland Smith's Expeditionary 
Troops staff was at Pearl Harbor, as 
were Admirals Spruance, Turner, and Con- 
oily and their staffs. The 77 th Division was 
still in the United States during the period 
when the initial plans for the landing were 
being worked out. 

On 29 March, General Geiger flew to 
Pearl Harbor, where for better than a 
week he conferred with General Smith and 
Admirals Turner and Conolly and their re- 
spective staffs. A week after Geiger's de- 
parture from Pearl Harbor, Admiral Con- 
olly flew to Guadalcanal, where the two 
commanders completed their planning and 
ironed out some of the many complicated 
problems involving naval-ground force co- 
ordination in the forthcoming landing. 7 

The upshot of these various conferences 
was the promulgation of one preferred and 
two alternate landing plans for Guam. The 
preferred plan called for simultaneous land- 
ings on the west coast by the 3d Marine 
Division between Adclup Point and Tatgua 
River and by the 1st Provisional Marine 
Brigade between the town of Agat and 
Bangi Point. The landing day (designated 

7 III Phib Corps Rpt Guam Opn, 3 Sep 44, 
pp. 1-2. 

W Day) was tentatively set as 18 June, 
three days after D Day for Saipan. 8 

To support the troops, Admiral Conolly 
split his Southern Attack Force into two 
groups, Northern and Southern Attack 
Groups. The former, commanded by Con- 
olly himself, was, under the preferred 
landing plan, to support the 3d Marine 
Division; the later, to be commanded by 
Rear Adm. Lawrence F. Reifsnider, USN, 
would perform the same function for the 
1st Provisional Brigade. The provisions for 
naval gunfire support during and after the 
landing closely paralleled those established 
for the Saipan operation. Before W Day, 
ships and aircraft of Admiral Conolly's 
Task Force 53 were to co-ordinate their 
bombardments with scheduled strikes by 
aircraft from Admiral Mitscher's Task 
Force 58. On W minus 2 and W minus 1, 
Task Force 53 was charged with responsi- 
bility for close-range support of underwater 
demolition teams and for destruction of 
coastal defense guns and antiaircraft and 
field artillery batteries on the landing 
beaches and of areas immediately inland. 
During the evenings ships of Task Force 
53 were to provide harassing fires and some 
time during this period were to conduct a 
diversionary bombardment on the east 
coast of Guam." 

On W Day itself the first priority for 
the fire support ships would be counter- 
battery fire, beginning at dawn, on known 
and suspected enemy positions. Secondary 
attention would be paid to local defenses 
by close-range fire. Third priority would be 
given to interdiction fire against any roads 

8 III Phib Corps Opn Plan 1-44, 11 May 44; 
Opn Plan 3-44, 30 May 44; Opn Plan 4-44., 9 
Jun 44. 

'■' III Phib Corps Opn Plan 1-44, Annex C, 
Naval Gunfire Support Plan. 



leading to the landing beaches. Shortly be- 
fore H Hour naval gunfire was to shift to 
close support fire on the flanks of the land- 
ing beaches. When the leading tractor 
waves were 1,200 yards from shore, cruisers 
were to lift their main batteries to the in- 
land areas with an accelerated rate of fire 
to neutralize mobile batteries and mortars. 
Five-inch gunfire was to be maintained 
along the beaches while naval aircraft 
strafed and bombed the same area. When 
the leading assault wave of LVT(A)'s was 
500 yards from shore, the 5-inch batteries 
were to shift fire to the near flanks of the 
beaches. One novel safety factor was intro- 
duced into the plans for Guam that had 
not been prescribed for the other landings 
in the Marianas. During the period when 
air and naval bombardment was to be con- 
ducted simultaneously, ships were ordered 
to restrict their fire to a maximum range 
of 8,000 yards, which meant in effect a 
maximum shell ordinate of about 1,200 
feet. At the same time pilots were in- 
structed to fly no lower than 1,500 feet. 10 

After H Hour, ships were to continue 
scheduled fires until ordered to stop. Call 
fires, it was planned, would be available as 
soon as communication with the shore fire 
control parties on the beaches was set up, 
and fire support ships were then to be pre- 
pared to deliver harassing fire, interdiction 
fire, star shell and searchlight illumination, 
and white phosphorus projectile fire on 
call. 11 

Admiral Conolly, as Commander, North- 
ern Attack Group, was to control the naval 
fire in support of the 3d Marine Division's 
landing over the Asan beaches, while Ad- 
miral Rcifsnider was to control bombard- 

10 TF 33 Opn Plan Ai62-44, 

Annex D, Vol. I, pp. 1-15. 

11 Ibid, 



ment of the brigade's beaches in the Agat 
area. Control of aircraft over both sets of 
beaches was to be exercised by Admiral 
Conolly alone through his Commander, 
Support Aircraft, stationed aboard the task 
force flagship, the AGC Appalachian. This 
officer was assigned control of combat air 
patrols, antisubmarine patrols, close air 
support of troops, and a variety of special 
missions. He would control not only the 
planes flown from the carriers attached to 
Task Force 53 but also planes flown from 
the fast carriers of Task Force 58 from the 
time of their arrival over the combat area 
until their departure for recovery by their 
parent ships. 12 

Before the arrival of Appalachian in the 
Guam area, an Advance Commander, Sup- 
port Aircraft, embarked on the cruiser 
Honolulu, would discharge these functions. 
A standby Commander, Support Aircraft, 
embarked in Admiral Reifsnider's flagship 
George Clymer, was assigned the temporary 
additional duty of Commander, Landing 
Force Support Aircraft. He was to assume 
this role under the command of General 
Geiger after the latter had established his 
command post ashore. The plan provided 
that when the Landing Force Support Air- 
craft commander was ready to take control 
of aircraft (land-based and carrier-based) 
in direct support of the troops, he was to 
advise Admiral Conolly and thereafter, un- 
der Conolly, would assume control of all 
support aircraft over Guam. Requests for 
carrier-based and distant land-based air- 
craft were to be sent by the Commander, 
Landing Force Support Aircraft, to Ad- 
miral Conolly, who would effect the ar- 
rangements for getting the planes on sta- 
tion and notify the Commander, Landing 

ay 44, 

12 III Phib Corps Opn Plan 1-44, Annex D, 
Air Support Plan. 



Force Support Aircraft, of their estimated 
time of arrival at the rendezvous point, 13 
This procedure would automatically give 
the landing force commander (General 
Geiger), through his air representative, 
more direct control over the aircraft cm- 
ployed in the support of his troops than 
was the case at Saipan and Tinian. There 
the Commander, Attack Force Support 
Aircraft, afloat, had kept close rein on all 
troop support missions flown from car- 

To avoid conflict between air support 
strikes and field artillery fire, Commander, 
Support Aircraft, or Commander, Landing 
Force Support Aircraft, was to request a 
cessation of fire from the commanding gen- 
eral of III Amphibious Corps Artillery for 
the duration of the air strike. Further co- 
ordination was to be secured through air 
observers, artillery spotters, and through 
the air co-ordinator. The function of the 
latter was to direct scheduled air strikes 
from the air and to report developments 
of the ground situation. Marine air ob- 
servers were to keep General Geiger in- 
formed on the ground situation, while 
artillery spotters would direct Marine ar- 
tillery fire. 

Preliminary air strikes on Guam were to 
begin on D Day at Saipan and last until 
W Day minus 1 under direction of the Ad- 
vance Commander, Support Aircraft. On 
W Day itself a major air strike was sched- 
uled to last for half an hour, from H Hour 
minus 90 minutes to H Hour minus 60 min- 
utes. During this period forty-six fighters 
and ninety-six dive bombers were to bomb 
and strafe gun positions and beach installa- 
tions in the two landing areas and sur- 
rounding territory. 14 

Change of Plans 

Following intensive amphibious training 
and rehearsals in the Guadalcanal area, 15 
the various Marine units of the III 
Amphibious Corps set sail aboard the 
transports and LST's of Task Force 53 and 
arrived at the staging area at Kwajalein 
Atoll on 8 June. After a brief period al- 
lowed for fueling, watering, and provision- 
ing, the convoy put to sea again and by 
15 June had arrived at its designated as- 
sembly area over a hundred miles to the 
east of Saipan. There it waited for ten days, 
cruising idly through the open seas, while 
higher authorities debated the feasibility of 
an early landing on Guam. 

Originally, W Day for the assault had 
been tentatively set as 18 June, but events 
on Saipan and in the adjacent waters made 
a postponement mandatory. As already in- 
dicated, the 27th Infantry Division, at 
first designated as reserve for the Saipan 
and Guam phases of the Marianas inva- 
sion, had to be committed in its entirety to 
Saipan, Furthermore, the Japanese Mobile 
Fleet had been sighted steaming toward 
the Marianas with the apparent intention 
of giving battle, and it was obvious folly to 
send the slow-moving troop transports and 
LST's of Task Force 53 into the waters 
west of Guam. Hence, Admiral Spruance 
canceled W Day and ordered Conolly's 
task force to remain out of danger well to 
the east of Saipan. 

By 25 June the situation ashore on Sai- 
pan had improved sufficiently to warrant 
releasing the HI Amphibious Corps from 
its duties as floating reserve for the V 

13 Ibid. 

14 Ibid., App. 1. 

ls Scc Major O. R. Lodge, USMC, The Re- 
capture of Guam, Historical Branch, G-3 Divi- 
sion, Headquarters, U.S. Marine Corps (Wash- 
ington, 1954) (hereafter cited as Lodge, Guam), 
pp. 26-28. 



Corps. Accordingly, the ships carrying the 
3d Marine Division sailed back to Eniwetok 
where they were followed five days later 
by the rest of the vessels carrying the 1st 
Provisional Marine Brigade. 1 " By early 
July, with the Saipan battle over two weeks 
old, thought could at last be given to de- 
ciding on a firm date for the landing on 
Guam. As a result of conferences among 
Admirals Turner, Hill, and Conolly and 
Generals Holland Smith and Geiger, 25 
July was recommended. These commanders 
deemed an earlier attack on Guam inadvis- 
able because, as Admiral Spruance ex- 
pressed it to Admiral Nimitz, "The char- 
acter of enemy resistance being encoun- 
tered in Saipan and the increase over the 
original estimates of enemy strength in 
Guam" made the presence of the entire 
77th Infantry Division necessary. 17 

Admiral Nimitz was anxious to schedule 
the assault for 15 July, by which time it 
was presumed that at least one regimental 
combat team of the 77 th Division could 
be dispatched to the scene of operations. 
Nevertheless, he deferred to the judgment 
of the officers present in the combat area 
and agreed to delay W Day until the whole 
of the Army division had arrived at Eni- 
wetok. On 6 July Admiral Spruance was 
advised that the last two regimental com- 
bat teams of the 77th Division to leave 
Hawaii could reach Eniwetok by 18 July, 
four days earlier than expected. Conse- 
quently, W Day was advanced to 21 July. 
The 305th Infantry of the 77th Division 
was constituted reserve for the 1st Pro- 
visional Brigade and ordered to land on the 
Agat beaches sometime after the brigade 
had gone ashore. The rest of the division 

Hi 1'F 53 Opn Rpt Guam, 10 Aug 44, pp. 5-6. 
17 Comdr Fifth Fleet, Final Rpt on Opn to 
Capture the Marianas, p, 5. 

was designated corps reserve and was or- 
dered to prepare to land "about William 
plus 2 Day on designated Beaches between 
Agat Village and Facpi Point" — also in 
the brigade zone. 18 

JJth Infantry Division 
Training and Preparation 

For the 77th Infantry Division, the in- 
vasion of Guam was to be the first chapter 
of a distinguished combat record in the 
Pacific war. Activated in March 1942, the 
unit spent its first two years in the United 
States undergoing training in basic infan- 
try warfare and in various specialties such 
as desert warfare at Camp Hydcr, Arizona, 
mountain warfare in West Virginia, and 
amphibious warfare in the Chesapeake Bay 
area. In May 1943 General Bruce assumed 
command of the division. A veteran of 
World War I, General Bruce was a gradu- 
ate of the Infantry School, the Field Ar- 
tillery School, the Command and Gen- 
eral Staff School, the Army War College, 
and the Naval War College. Before he as- 
sumed command of the 77th Division he 
completed a tour of duty in the Operations 
and Training Division of the War Depart- 
ment General Staff and commanded the 
Tank Destroyer Center at Camp Hood. 19 

By March 1944 the division was located 
at Oahu, and for the next three months it 
was put through an intensive indoctrina- 
tion in the techniques of warfare peculiar 
to the Pacific area under the direction of 
General Richardson's United States Army 

18 III Phib Corps Opn Plan 1-44, Addendum, 
10 Jul 44. 

19 Ours To Hold It High, The History of the 
77th Infantry Division in World War II By Men 
Who Were There (Washington: Infantry Journal 
Press, 1947), pp. 1-40. 



Forces in the Central Pacific Area ( USAF- 
ICPA). Infantrymen were trained as flame 
thrower and demolition men so as to avoid 
the necessity of relying exclusively on en- 
gineers to perform these functions in com- 
bat. Officers and noncommissioned officers 
took a forward-observers course to reduce 
dependence on artillery personnel. The 
706th Tank Battalion trained with the in- 
fantry regiments in mutual close-in sup- 
port, combined maneuver problems, and 
landing operations. The entire division 
spent six days at the Unit Jungle Training 
Center on Oahu. Amphibious training for 
the infantrymen consisted of net-climbing, 
embarkation, and debarkation from mock- 
up ships, transfer of personnel and equip- 
ment from LCVP's to LVT's at sea, and 
landing on beaches in wave formation. Ar- 
tillery units conducted test landings from 
LST's and practiced landing operations in 
DUKW's and LVT's with battalion landing 
teams. Experiments in loading and landing 
155-mm. guns from LCM's, LST's, and 
LCT's were made. The 77th Division Re- 
connaissance Troop held four days of 
practice with destroyer escorts, and part 
of the 2o,2d JASCO trained with Navy 
aircraft at Maui. The JASCO's shore fire 
patrol party conducted destroyer firing ex- 
ercises at the naval gunnery range on Ka- 
hoolawe Island. The only important feature 
missing from the program was the cus- 
tomary last-minute ship-to-shore rehearsal, 
which had to be foregone because of the 
lack of time. For most of the period the 
XXIV Corps, to which the division had 
initially been assigned, assisted in the 
training. Not until 22 June, almost on the 
eve of its departure from Oahu, was the 
division released to the V Amphibious 
Corps, the Marine Corps' administrative 

and training command in the Hawaiian 
area. 20 

At the time the division set sail from 
Hawaii in the first week of July, it was 
still a matter of doubt as to how and where 
it would be employed on Guam. Not until 
the middle of the month when the troop 
transports had reached Rniwetok was Gen- 
eral Bruce fully apprised of the intentions 
of his superiors in. regard to his unit. This 
meant that there was very little time before 
the target date to complete plans and to 
disseminate them to subordinate units. An 
additional handicap was the fact that not 
until they arrived at Guam itself were the 
commanding general and his staff able to 
establish personal contact with higher, ad- 
jacent, and supporting units. 21 

Despite these difficulties, by 15 July 
General Bruce was able to promulgate 
three plans (one preferred and two alter- 
nates) for the division's commitment. The 
plan already devised by III Amphibious 
Corps had contemplated the seizure of a 
Force Beachhead Line to extend from 
Adelup Point along the Mount Alutom- 
Mount Tenjo-Mount Alifan ridge line to 
Facpi Point. According to the corps' pre- 
ferred plan, the 3d Marine Division would 
land between Adelup Point and the mouth 
of the Tatgua River and move south to 
the Apra Harbor area while the 1st Pro- 
visional Marine Brigade, with the 305th 
Regimental Combat Team of the 77th Di- 
vision attached, would land between the 
town Agat and Bangi Point and then wheel 
north to the base of Orote Peninsula. 

29 USAFIGPA Participation Rpt Marianas, 
Vol. I, pp. 24, 68, 171, 172, 266, 276, 289-93; 
Vol. II, pp. 315, 337. See Appendix D for order 
of battle of the 77th Infantry Division for the 
Guam Operation. 

21 77th Inf Div Opn Rpt Forager, 21 Jul-16 
Aug 44, p. 17. 



Gearing his own plans to this schedule 
of operations, General Bruce directed the 
2d Battalion, 305th, with a platoon of the 
706th Tank Battalion, to assemble at the 
line of departure two hours after H Hour 
and be ready to land on order of the 1st 
Provisional Marine Brigade. The other two 
battalions of the 305th were free of specific 
instructions except to be ready to debark 
and land, also on brigade order. As corps 
reserve, the 306th and 307th Regiments 
would land on corps order over the same 
beaches as the brigade, move to assembly 
areas, and be prepared to relieve the bri- 
gade of the duty of defending the final 
beachhead line. 22 

The two alternate plans prepared for the 
Army division's commitment were both 
based on the assumption that the Marines' 
preferred plan would be put in effect. One 
of these contemplated a landing by the 
306th and 307th Infantry Regiments near 
Adelup Point, whence they would move 
southwest to assembly areas and be pre- 
pared to attack either south toward Mount 
Tenjo or southeast toward Pago Bay. The 
second called for landings by the same two 
regiments on the northwestern coast be- 
tween Uruno and Ritidian Points, from 
which positions they would move southwest 
in order to secure from the rear a beach- 
head at Tumon Bay. 23 

This second plan was particularly dear 
to General Bruce's heart. 24 He wanted the 

two regiments of the 77th Division that 
were in corps reserve (the 306th and 
307th) to land at dawn or just before 
dawn near the northwest tip of Guam 
about four days after the initial amphib- 
ious assault. They would then drive rapidly 
south and capture a beachhead at Tu- 
mon Bay from the rear. They would land 
without heavy equipment, but once the 
Tumon Bay beaches were secured neces- 
sary supplies and equipment and possibly 
other infantry elements could be landed 
there. The concept of the plan, as Bruce 
expressed it, was "for the 77th Division to 
become a hammer striking forwards and 
eventually on the anvil, i.e., the Force 
Beach Line. Should the enemy divert suf- 
ficient forces to halt this Division for any 
appreciable length of time it should be 
possible for the 77th Division to become 
the anvil and the forces occupying the FBL 
to become the hammer," 2G 

Immediately upon his arrival at Eniwe- 
tok on 1 1 July, General Bruce, with char- 
acteristic enthusiasm, pressed this scheme 
on the Marine commanders present. None 
of them warmed to the proposal, and Maj. 
Gen. Allen H. Turnagc told him he had 
better drop the idea. The Marines were 
reluctant to divert the corps reserve to a 
secondary landing for fear it might be 
needed to support the assault troops at 
the main beachheads. Undaunted by this 

32 77th Inf Div Opn Plan I, if, Jul 44, 77th 
Inf Div Initial Opn Plans and A dm Orders 
Guam, Jun— Jul 44, 

33 77th Inf Div Opn Plan II and III, 15 Jul 
44, 77th Inf Div Initial Opn Plans and Adm 
Orders Guam, Jun-Jul 44. 

114 Throughout his entire World War II service 
as commanding general of the 77th Division, Gen- 
eral Bruce was an enthusiastic supporter of the 
concept of the "amphibious end run," that is, of 
secondary amphibious landings to the rear of the 

main Force Beachhead Line. During the Lcyte 
operation he succeeded in getting the idea accepted 
with the result that the 77th Division made a 
highly successful landing at Ormoc. At Okinawa 
he tried to persuade his immediate superiors to 
let his division make a similar landing on the 
southern coast of the island to the rear of the 
Japanese main line of resistance, but the plan was 

35 Ilq 77th Inf Div, Alternate Opn Plan Em- 
ployment 77th Div in Corps Reserve, 14 Jul 44. 



cold reception, Bruce sent a despatch out- 
lining his plan to General Geiger, who had 
already sailed for the Marianas aboard Ad- 
miral Conolly's flagship, Appalachian. The 
corps commander rejected the plan on the 
grounds that it was then too late to make 
any radical changes. 26 

Loading and Embarkation 
yyth Division 

In the Hawaiian area where the 77th 
Division loaded and embarked for the 
Guam operation, its logistical needs were 
handled by the supply section of General 
Richardson's headquarters and its subordi- 
nate agencies and by Commander, Service 
Force Pacific, a naval organization, Hol- 
land Smith's staff determined the amounts 
of each class of supplies to be landed 
initially and supervised or at least checked 
the tentative loading plans. Tn the early 
phases of the logistical planning, it was not 
considered necessary to provide initial com- 
bat supplies to the division because the 
earlier operation plans did not call for its 
commitment within thirty-five days of the 
Marine assault landings. 27 Once the divi- 
sion was designated area reserve and then, 
in mid- June, alerted for movement to the 
Marianas, supply activities were naturally 

General Bruce's first supply order was 
issued on 24 June and specified the levels 
of initial supply for each class: 

Class I 

Type B 10 days 

10-in-i pack jo days 

Type C 7 days 

Type K 3 days 

Type D a days 

Assault candy ration 1 per man 

Rations accessory' pack 20 days 


In 5-gallon containers and 55-gallon drums, s 
gallons per man per day for 5 days; 1 water 
point unit for each RGT; 2 water distillation 
units for each engineer combat battalion. 

Class II 

20 days maintenance of clothing, equipment, 
and general supplies, bulk clothing and individual 
equipment carried by RCT's to be equally dis- 
tributed in all ships and landed early. 



Fuels and lubricants 

20 days 

Class IV 

Chemical Warfare 

20 days 


20 days 


20 days 


20 days 


30 days 


20 days 

Class V 

Antiaircraft weapons 10 CINCPOA U/F 
All other weapons 7 CINCPOA U/F 2R 

Like the 27th Division before it went to 
Saipan, and unlike any of the Marine 
divisions destined for the Marianas, the 
77th complied enthusiastically with General 
Holland Smith's directive that between 25 
and 50 percent of all supplies be palle- 

36 Ltr, Lt Gen Andrew D. Bruce, USA (Ret.), 
to Gen A. C. Smith, i i Feb 55, with Incl 2 (copy 
of Forager Notes kept by General Bruce), OCMH. 

37 TF 56 Rpt Forager, Incl E, G-4 Rpt, p. 8. 

*' H 77th Inf Div Administrative Order 1, 24 
Jun 44, pp. 1-3. Later the number of days supply 
of 10-in-i pack rations was changed to 20. Maj. 
Gen. Andrew D. Bruce, "Administration, Supply, 
and Evacuation of the 77th Infantry Division on 
Guam," Military Review, XXIV (December, 
1944), 3. For the composition of CINCPOA units 
of fire,|sce Appendix B. I 



tized. 29 All together, the division built 
about five thousand pallets, but about a 
thousand of these were dismantled before 
being loaded aboard ship because they 
would not fit into available spaces in the 
holds or because they were too difficult to 
handle in holds where fingerlifts were not 
available. 30 In contrast, the marines bound 
for Guam palletized none of their supplies, 
partly because they lacked the building ma- 
terials and equipment to handle them and 
partly because they were still skeptical as 
to whether pallets could be satisfactorily 
hauled ashore over a coral reef. 31 

To lift the more than 18,000 troops and 
the 21,428 tons of supplies and equipment 
of the 77th Division from Hawaii to Eni- 
wetok and then on to Guam, the Navy 
provided seven attack transports, four 
transports, three AKA's, two AK's, and 
three LST's. The LST's carried 6 1 2 troops 
and the 53 DUKW's allotted to the divi- 
sion. The DUKW's were the only amphib- 
ian vehicles allowed the 77th; no LVT's 
were taken along since the division was not 
scheduled to go into the beach in assault. 32 

Loading the vessels was complicated by 
the fact that the 77th Division had less 
than two weeks' advance notice as to how 
many ships would be made available to it 
and what their characteristics would be. 
Loading plans therefore had to be sketchy 
and tentative, since they could not be made 
final until approved by the commanding 
officer of each ship. The vessels themselves 

' 2a TF 56 Rpt Forager, Incl G, Transport QM 
Rpt, p, 15. 

so 77th Inf Div Adm Order 1, 24 Jun 44; 
Bruce, "Administration, Supply, and Evacuation 
. . .," Military Review, XXIV, 6. 

31 III Phib Corps Rpt Guam, Incl A, Supply 

32 TF 56 Rpt Forager, Incl G, Transport QM 
Rpt, pp. 315, 317, InclBB. 

did not arrive at Oahu until forty-eight 
hours before the date set to begin loading. 
To troops who were about to embark on 
their first amphibious operation, the delay 
of course was maddening. One battalion 
commander later recalled that his transport 
quartermaster, the troop officer in charge 
of loading, "worried for 5 straight days 
without sleep, as did most of his assis- 
tants." :w The 305th Regimental Combat 
Team was the first of the regiments to be 
loaded, and it left Honolulu on 1 July. 
Embarkation was hasty, troops and cargo 
were loaded simultaneously, and much con- 
fusion resulted. The other two regiments 
had to await the return of transports that 
had been involved in the first phase of the 
Marianas operation at Saipan. They did 
not leave Hawaii until 8 July, 34 

The Marines 

The Marine units embarking for the 
Guam operation had one distinct advan- 
tage over the 77th Division in that their 
assigned shipping was present in their 
mounting area, the Solomon Islands, well 
in advance of the embarkation date. Ad- 
miral Conolly and his staff arrived at 
Guadalcanal on 15 April for a stay of 
nearly a month of preliminary planning 
with the III Amphibious Corps. The naval 
forces assigned to Task Force 53 were 
largely from Admiral Halscy's South Pa- 
cific Area and had engaged in the Hollan- 
dia operation before putting in to bases in 
the Solomons on 10 May. Before their re- 
turn for attachment to Conolly's task force, 
arrangements had been made to station 

33 Ltr, Col James E. Landrum, Jr., to Gen A. C. 
Smith, 27 Apr 55, and Incl, OCMH. 

34 Ibid., p. 20; 77th Inf Div Opn Rpt Forager, 
305th RCT Rpt, p. 12. 



them at Guadalcanal, Efate, Espiritu Santo, 
and Hathorn Sound so as to avoid over- 
crowding Guadalcanal and Tulagi. 35 

Because of the presence of the ships in 
the immediate area of the points of em- 
barkation, there was ample time for the 
transport division commanders and ship 
captains to check thoroughly the loading 
plans of the various units. Such difficulty 
as was encountered centered primarily 
around the loading and embarkation of the 
ist Provisional Marine Brigade, which had 
only recently been formed and had not had 
as much time to prepare detailed plans as 
had the 3d Marine Division. 36 

More important was the fact that early 
in May the naval lift allotted to the assault 
units was unexpectedly ordered to carry 
more than five hundred troops of Maj. 
Gen. Henry L. Larscn's Island Command 
Headquarters Group, which was destined 
for garrison duties on Guam after it was 
secured. This raised additional demands on 
the already limited shipping space, and the 
result was that the marines had to leave a 
good number of their organic vehicles be- 
hind when they sailed from the Solomons. 
Their supply of amphibian vehicles, how- 
ever, was not curtailed. The 358 LVT's of 
the 3d and 4th Amphibian Tractor Bat- 
talions were assigned respectively to the 3d 
Marine Division and the ist Provisional 
Marine Brigade. Their job, after delivering 
the assault troops ashore, was to transship 
cargo and personnel from landing craft over 
the reef and thence to shore. In addition 
to these vehicles, the marines were sup- 
plied with a hundred DUKW's. 37 

35 TF 53 Opn Rpt Guam, pp. 1-2. 

36 III Phib Corps Rpt Guam, Incl A, Supply 
Rpt, pp. 1-3. 

37 Lodge, Guam, pp. 21-22. See Appendix E 
for Order of Battle of III Amphibious Corps (less 
77th Division) . 

Preliminary Bombardment 
Naval Gunfire 

No matter what the immediate incon- 
venience to American forces caused by Ad- 
miral Spruance's postponement of the 
scheduled landing day on Guam, the long- 
run consequences of that decision were for- 
tunate. The postponement of W Day from 
18 June to 21 July made possible a more 
prolonged preliminary air and sea bom- 
bardment against Guam than against any 
other island in the Pacific during the war. 
The marines of the III Amphibious Corps 
who had chafed and fretted at being con- 
fined to their ships in the sweltering lagoon 
of Eniwetok later had good reason to be 
thankful for their enforced inactivity. 

The first American naval shells to hit 
Guam were fired from ships of a small task 
group from Task Force 53 on 16 June, the 
day after the landing on Saipan. For an 
hour and three quarters, the cruiser Hono- 
lulu, the battleships Pennsylvania and 
Idaho, and several destroyers, all supported 
by planes from accompanying aircraft car- 
riers, bombarded the west coast of the is- 
land. The damage done appears to have 
been negligible, but the raid did alert the 
Japanese as to the probable American 
choice of landing beaches in the forthcom- 
ing invasion, 38 

The enemy ashore mistook the shelling 
for the usual last-minute preliminary to an 
assault landing, and one Army lieutenant 
wrote in his diary : 

For the first time I saw the enemy fleet and 
was under its gunfire, I regret very much that 
we arc powerless to do anything but to look 

44, p. 29 ; Morison, New Guinea, and the Mari- 
anas, p, 377. 



at the enemy which came in barely 10,000 
meters away. They shelled us steadily for two 
hours. Our positions were hit fourteen times. 
Fortunately none was injured, . . , Wc think 
that at last the enemy will land tonight, and 
so we will observe strict alert all night. We 
were issued hand grenades and are now wait- 
ing for the enemy to come. 39 

By next morning, of course, the Ameri- 
can ships had disappeared over the horizon, 
much to the disappointment of the lieu- 
tenant, and probably of most of his com- 
rades. Impatiently, he wrote, "If the en- 
emy is coming, let him come. The spirit to 
fight to the death is high. We are anxiously 
waiting but nothing unusual has happened 
so far as dawn breaks." 40 

The next surface ship strike against 
Guam occurred on 27 June when a 
small detachment of cruisers and destroyers 
(Task Unit 58.4.5) from Admiral Mits- 
cher's carrier fleet made a quick run into 
the waters off Guam and Rota, sank a 
small harbor tug and two barges in Apra 
Harbor, and set fire to some oil storage 
tanks ashore. Three days later Destroyer 
Division 46 shelled the airfields on Orote 
Peninsula. 41 

Then, on 8 July, began the greatest 
single naval bombardment program of the 
war — greatest at least in terms of time ex- 
pended. For thirteen days the Japanese gar- 
rison on Guam was treated to the most 
spectacular display of shore bombardment 
that the U.S. Navy had yet produced. 

First to arrive were four heavy cruisers, 
twelve destroyers, and two escort carriers 
of Task Group 53.18 commanded by Rear 

ss CINCPAC-CINCPOA Trans 10634, ex- 
tracts from the diary of 2d Lt Imanishi, Rai 32 11 
To (38th Inf Re^t, agth Div) (hereafter cited as 
Irnanishi Diary), 

40 Ibid. 

44, p. 29. 

Adm. C, Turner Joy. The group's primary 
mission was to destroy coastal defense and 
heavy antiaircraft guns. Secondary targets 
were warehouses, command posts, commu- 
nications facilities, and troop concentra- 
tions. Co-ordinating with the planes from 
two task groups of Task Force 58 that ar- 
rived in the area about the same time, the 
cruisers and destroyers were responsible for 
one half of the island while the planes 
bombarded the remainder. At noon each 
day the two exchanged areas of responsibil- 
ity. Meanwhile, planes from the two escort 
carriers flew combat air and antisubmarine 
patrol. At night each warship delivered 
harassing fire against the island. On 1 2 
July the battleships New Mexico, Idaho, 
and Pennsylvania arrived to add their bit 
to the fireworks. Two days later Admiral 
Conolly himself put in. his appearance 
aboard the AGC Appalachian and there- 
after personally took charge of co- 
ordinating all naval and air bombardment. 
The same day, the battleship Colorado 
joined the bombardment force, as did 
California and Tennessee on the 19th. By 
the time the marines arrived to invade the 
island, a total of six battleships and nine 
cruisers with their escorting destroyers 
were saturating Guam with naval shells of 
all varieties. For this period of thirteen days 
(exclusive of W Day itself) naval ammu- 
nition expenditures against shore targets 
totaled 836 rounds of 1 6-inch, 5,422 of 
14-inch, 3,862 of 8-inch, 2,430 of 6-inch, 
and 16,214 of 5-inch shells. 42 

At the invasion of Roi-Namur Admiral 
Conolly had earned the sobriquet "Close- 
in Conolly" for his insistence that warships 

42 TF 53 Opn Rpt Guam, Incl E, Intel Rpt, 
p. 5; TF 56 Rpt Forager, Incl G, Naval Gunfire 
Support Rpt, p. 71; CINCPAC-CINCPOA Opns 
in POA, Jul 44, p. 29. 



cruise close to shore when firing at land 
targets. 43 At Guam, he reaffirmed his right 
to the title, but more important was the 
systematic procedure he introduced for co- 
ordinating naval gunfire and aerial bom- 
bardment and checking the results of each. 
A target board of six officers, representing 
the air, gunnery, and intelligence sections 
of the staff, was set up to assign primary 
missions for air strikes and naval gunfire 
and assess the damages daily before desig- 
nating the next day's targets. Aerial photo- 
graphs were taken each morning and on 
the basis of these damage was assessed and 
new targets were assigned. In these opera- 
tions, the admiral's staff was aided by the 
presence aboard Appalachian of General 
Geiger who, as commanding general of the 
landing force, naturally had the greatest 
personal concern about the accuracy both 
of the bombardment and of the damage re- 
ports submitted afterward, 44 

During the later stages of the prelimi- 
nary bombardment, one additional duty was 
imposed on the ships present— that of sup- 
porting naval underwater demolition teams. 
All together, three teams were made avail- 
able for the Guam landings. The procedure 
was one that had by now become stand- 
ardized in the Pacific. Swimmers dis- 
embarked from their mother APD into 
LCPR's that took them close to the reef 
before putting them in the water to swim 
in the rest of the way to inspect the reef 
itself. In the meantime, four LCI gunboats 
lay to just off the reef and fired their 40- 
mm. and 20-mm. guns over the heads of 
the swimmers. On each flank of the 

13 See Growl and Love, Gilberts and Marshalls, 

4i TF 53 Opn Rpt Guam, Incl B, Naval Gun- 
fire Support Rpt; Lodge, Guam, p. 34; Morison, 
New Guinea and the Marianas, pp. 380-81, 

LCI(G)'s was a destroyer firing five-inch 
shells farther inland, while the APD fol- 
lowed astern of the line of gunboats, also 
firing. After the small boats had picked up 
their swimmers, the covering ships con- 
tinued their fire on the beaches in an effort 
to interdict the area where the enemy was 
attempting to make repairs. 

On 14 July Underwater Demolition 
Team 3, aboard the APD Dicker son, ar- 
rived in the area and for three days con- 
ducted reconnaissance of the chosen land- 
ing beaches and other segments of the west- 
ern coast line. On 17 July Underwater 
Demolition Teams 4 and 6 put in their ap- 
pearance. Actual demolition work began 
that evening. The obstacles discovered on 
the Agat beaches were chiefly palm log 
cribs filled with coral and connected by 
wire cable. On the Asan beaches wire cages 
filled with cemented coral were spaced 
about every five feet. Only occasional strips 
of barbed wire were found, and no under- 
water mines. All together, 640 obstacles 
were blown-up off Asan and about 300 off 
Agat by hand-placed demolitions. 45 Some 
of these, at least, had been constructed as 
recently as 3 July, by which time the Jap- 
anese had been tipped off as to the prob- 
able landing beaches to be used by the 
invaders. 46 

Aerial Bombardment 

In preparing the way for the amphibious 
assault on Guam, four main duties fell to 
the air arms of the Army and Navy. They 
were to neutralize Truk and the other is- 
lands in the Caroline group from which 
the Japanese might be expected to send 

45 TF 53 Opn Rpt Guam, p. 10; CINCPAC- 
CINCFOA Opns in POA, Jul 4.4, pp. 31-43. 
1,6 Imanishi Diary. 



their own aerial strength into the southern 
Marianas, prevent intervention by Japa- 
nese carrier-borne planes, photograph the 
island, and soften the target itself with an 
accelerated program of aerial bombing and 

Starting in mid-March and continuing 
even after Guam had been secured, Army 
Air Forces bombers of the Seventh, Thir- 
teenth, and, later, Fifth Air Forces con- 
ducted a series of devastating raids against 
the Carolines, chiefly Truk and Woleai. On 
one occasion, during the last two days of 
April, they were joined by Admiral Mits- 
cher's Task Force 58, which dropped 748 
tons of bombs on Truk while retiring from 
the Hollandia invasion. The major credit 
for keeping the Carolines neutralized, how- 
ever, fell to the Army Air Forces. By the 
time of the invasion of the southern Mari- 
anas, the island of Truk, once the leading 
Japanese bastion in the Central Pacific, had 
been rendered virtually useless. 47 A like 
fate had befallen the great Japanese Mo- 
bile Fleet at the hands of Task Force 58 
during the Battle of the Philippine Sea. By 
20 June it was clear that the invaders of 
Guam need have no fear of serious Japa- 
nese threats from the air. 

To Task Force 59, commanded by Maj. 
Gen, Willis H. Hale, AUS, fell the chief re- 
sponsibility for aerial photographic recon- 
naissance of the Marianas. Seventh Air 
Force and shore-based Navy bombers, both 
under General Hale's command, main- 
tained armed reconnaissance over all the 
southern Marianas for more than two 
months before the first landings on Saipan. 
The first mission over Guam was carried 
out on 6 May by ten Army B-24's escort- 

ing six Navy PB4Y's. Five of the planes 
were shot down over the target by enemy 
fighters; six others were damaged. Again 
on 24 May, 29 May, and 6 June flights of 
B-24's and PB4YS made the trip over 
Guam, taking photographs and dropping 
token loads of bombs on targets of oppor- 
tunity. 48 Of the 6 June raid, a Lieutenant 
Imanishi wrote despairingly, "There were 
9 B-24's, 49 but not one of our planes 
went up to meet them. We felt dis- 
heartened. Just how desirous our air force 
is of fighting is open to doubt." 50 

Desirous of fighting or not, the Japanese 
pilots stationed on Guam were soon to lose 
the means of doing so. Shortly after the 
photographic flight on 6 June, Admiral 
Mitscher's fleet showed up to begin its 
methodical destruction of enemy aircraft 
and air facilities. In the belief that the is- 
land would be invaded on 18 June, Com- 
mander, Task Force 58, first unleashed his 
mighty armada of fighters and bombers 
against Guam and nearby Rota on 11- 12 
June. In the ensuing air battle, a total of 
150 Japanese planes was reported de- 
stroyed in the air or on the ground. For the 
next four days, one or more of Mitscher's 
task groups carried out strikes against air- 
craft facilities, runways, coastal guns, and 
antiaircraft positions on Guam and Rota. 51 

Against this overwhelming naval air- 
power, the Japanese were almost helpless. 
Wrote Lieutenant Imanishi, "It is espe- 
cially pitiful that we cannot control the air. 
We can only clench our fists with anger and 

47 See above, pp. 71-72; Craven and Cate, AAF 
IV, pp. 676-90; USSES, The Reduction of 

48 Army Air Forces in the Marianas Cam- 
paign, Operation ForageRj Mar-Aug 44, pp, 
7 -io, MSinOCMH. 

la Actually there were seven R-24's and four 

r>0 Imanishi Diary. 

44, pp. 69-93. 



watch," r ' 2 At the same time, a Japanese 
private noted that he and his companions 
were unable to leave their shelters and help 
repair the damage because of the bomb- 
ings. 51 Another enlisted man complained, 
"The number of enemy planes was said to 
be more than 500 today, while not one of 
our planes took to the air. I felt a bitter 
resentment at the manner in which the en- 
emy stressed his air power," 54 

Yet there was still some fight left in the 
Japanese air contingent on Guam, for on 
the evening of 15 June a few planes took 
off from Orote field to launch a low-level 
torpedo attack against the American car- 
riers offshore. As a result, two of Mitscher's 
task groups next day concentrated heavily 
on Guam to prevent a repetition of the 
previous evening's attacks. 55 During the 
two-day Battle of the Philippine Sea, the 
fields of Guam again received the attention 
of Mitscher's fliers. Japanese land-based 
planes still undamaged by previous raids, 
as well as carrier planes that had flown in 
from Ozawa's fleet, constituted a threat on 
Mitscher's flank and rear that could not be 
overlooked. On the morning of 19 June, be- 
fore the Great Marianas Turkey Shoot had 
even gotten well under way, two separate 
air battles were fought over Guam, both 
ending in victory for the Americans. Even 
during the course of the main battle it- 
self, which was fought well out to sea, Mits- 
cher kept one contingent of fighters and 
bombers over Guam to interdict the air- 

S2 Imanishi Diary. 

83 CINGPAC-GINCPOA Trans 10996, ex- 
tracts from the diary of Leading Pvt Murano, 
Koko (2d Bn, 10th IMR). 

S4 CINCPAC-CINCPOA Trans 10802, ex- 
tracts from the diary of Cpl Susuki, Tai ( Yoihi- 
kawa Unit). 

44, Annex A, pp. 8r-g8. 

fields and prevent any remaining planes 
from taking off to join Ozawa's carrier 
planes. All together, about fifty Guam- 
based planes were destroyed on the 19th 
alone, and the fields themselves were at 
least temporarily put out of business/' 6 
That night, when about fifteen Japanese 
carrier bombers attempted to make emer- 
gency landings there, they found the fields 
too torn up to do so and, being out of 
fuel, had to crash. 57 

The raids of ig June all but delivered 
the coup de grace to Japanese airpower 
on Guam. Occasionally, in the weeks that 
followed, a few Japanese planes flew into 
Orote from Yap and other islands in the 
Carolines, but they posed no real threat. 
On 4 July one of Mitscher's task groups 
(Task Group 58.3) returned to conduct 
a daylight raid over the island, and from 
6 through 17 July two other carrier 
groups (Task Groups 58.1 and 58.2) alter- 
nated daily in strikes over Guam and Rota. 
Primary targets were coastal and antiair- 
craft guns, supply dumps, airfield installa- 
tions, and the towns of Asan, Piti, and 
Agat. These strikes were co-ordinated with 
those of the escort carrier planes and the 
naval bombardment ships of Admiral Con- 
olly's Task Force 53- HS On the last three 
days before the landing, the Japanese on 
Guam witnessed the full weight of Ameri- 
can naval airpower in a mounting cre- 
scendo of aerial fury. On 18 July planes 
from the two task forces flew 662 bombing 
sorties and 31 1 strafing attacks, on the 19th 
the number increased to 874 and 392, and 
on the day before the landings to 1,430 
and 614. The total tonnage of bombs, 

■ r,B Morison, New Guinea and the Marianas 
pp. 257-63, 274-77. 
a1 Imanishi Diary. 
KS See above! pp. 320-21 J 



depth charges, and rockets dropped and 
launched during these three days came to 


Assessment of Damage 

As night closed in on 20 July, it seemed 
impossible to those aboard the flagship 
Appalachian that the Japanese on Guam 
could put up anything but token resistance 
to the troops that would go in the next day 
in amphibious assault. Maj. William M. 
Gilliam, USMC, who was Gciger's naval 
gunfire officer, reported, "When the morn- 
ing of the landing arrived, it was known 
that the assault troops would meet little 
resistance." no Admiral Conolly's staff be- 
lieved, "Not one fixed gun was left in 
commission on the west coast that was of 
greater size than a machine gun." fi! 

These conclusions were to prove some- 
what extravagant, as the marines next day 
discovered to their sorrow. Testimony given 
after the war by Lt. Col. Hideyuki Takeda, 
IJA, who was a staff member of the 'jgth 
Division, provides a corrective to the Amer- 
ican reports on which these optimistic con- 
clusions were based. 

Conventional construction, Takeda re- 
ported, consisting of buildings reinforced 
on an emergency basis, was completely 
destroyed when it received direct hits. 
Field positions that were hit by shells were 
completely destroyed, and of those on or 
near the landing beaches, over 50 percent 
were demolished. Half-permanent positions 
in which the hard agent cascajo (a type 
of coral ) was used and that were reinforced 

!»< p fF 153 Opn Rpt Guam, Incl C, App. 5. 

0,0 III Phib Corps Rpt Guam, Incl G, Naval 
Gunfire Support, p. 3, 

(il TF 5 3 Opn Rpt Guam, Incl B, Naval Gun- 
fire Support, p. II. 

with concrete about 50-cm. thick remained 
in good condition except in cases of direct 
hits. Those that were hit by shells were 
more than 50 percent destroyed. Perma- 
nent positions with concrete over one meter 
thick remained in perfect condition even 
after receiving direct hits. All open naval 
gun emplacements were completely de- 
stroyed before the landings. Of those naval 
guns emplaced in caves, about half re- 
mained operational at the time of the 
landings, but they were soon put out of 
commission by counterbattery fire that 
closed up the cave mouths where they were 
located. Antiaircraft artillery on the island 
sustained damage from naval gunfire only 
once, and so long as Japanese antiaircraft 
ammunition lasted the Japanese were rea- 
sonably safe from American planes. Harbor 
installations received almost no damage, 
water pipes received only one direct hit, 
and power installations were all located in 
caves and so escaped damage. Most mili- 
tary boats were sunk. Naval gunfire had 
no effect against construction in the valleys 
or in the jungle and had very little effect 
against the interior parts of the island over 
two and a half miles from the shore line. 
American airpowcr, reported Colonel 
Takeda, succeeded in knocking out the air- 
fields on Guam but posed little threat to 
defense positions because there was little 
bombing of Japanese gun emplacements 
from the air. By far the most important 
effect of aerial bombing and strafing was 
the extreme limitation it placed on Japa- 
nese ground movement during daylight 
hours. However, neither naval guns nor 
aircraft succeeded in causing any serious 
interruption in communications on Guam. 
Takeda could not remember a single case 
where telephone lines were cut because of 
naval gunfire. As of 2 1 July, Headquarters, 



2gth Division, in command of the defense 
of Guam, possessed perfect wire and wire- 
less communications with the 18th Infan- 
try Regiment, the gStk Infantry Regi- 
ment, the force on Orote Peninsula, forces 
south of Pago on the east coast, and forces 
at Tarague on the north tip of the island. 
Perfect field telephone communication with 
the 4.8th Independent Mixed Brigade was 
maintained. Headquarters also had un- 
interrupted wireless communication with 
Rota, as well as with Imperial General 
Headquarters in Tokyo." 2 

In spite of the limited effectiveness 
of American preliminary bombardment, 
Takeda's testimony does indicate that it 
produced certain substantial results. Many 
of the buildings that were destroyed by di- 
rect hits, such as hospitals, warehouses, 
and office buildings in the towns of Piti, 
Agana, and Agat, housed military person- 
nel and equipment. Takeda's own appreci- 
ation of the important role played by 
American air and sea power in reducing 
the defenses of Guam emerges through the 
crude but clear translation of the closing 
words of his postbattle report : 

Among the battle colored by the holy 
blood of the dead I can find out the only 
lesson: The powerful air and sea powers 
make ground forces to defend island unneces- 
sary. That is, the defence of island depend- 
ing merely upon the isolated and helpless 
ground forces cannot be existed in the world. 
If the defence depending only upon the 
ground forces succeeded it would only be 
clue to the fact that the island was neutral- 

* ia Lt Col Hideyuki Takeda, I J A, Outline of 
Japanese Defense Plan and Battle of Guam Is- 
land, translated by Major Sato, IJA, Incl to Ltr, 
Col W. S. Coleman to Comdt USMC, 4 Oct 46; 
Lt Col Hideyuki Takeda, Ltr to Brig Gen J. 
C. McQueen, USMC, Dir Marine Corps Hist, 20 
Feb 52, translated by Thomas G. Wilds, OGMH 
(both in Records Sec Hist Br G-3, Hq USMC). 

ized, troops on it would hardly exist and they 
could perform their duty to defend the is- 
land because the enemy did not land on it. 63 

Intelligence of the Enemy 

Considering the fact that Guam had 
been a U.S. possession for more than forty 
years, American intelligence of the island's 
road system and terrain was remarkably 
incomplete. The War Department in June 

1943 had prepared and published a gen- 
eral survey of the island, and in February 

1 944 the Office of Naval Intelligence circu- 
lated a voluminous bulletin containing all 
kinds of information about hydrographic 
conditions, ground contours, road systems, 
weather, and the native population/' 4 
Neither of these studies was any further up 
to date than 10 December 1941, the date 
of the Japanese occupation, nor could the 
information supplied by American service- 
men and native Guamanians who had lived 
on the island before the occupation give 
the planners any idea of Japanese defense 
installations or dispositions. General Geiger 
asked permission to send in small patrols 
by submarine to contact natives and "see 
behind the curtain," but the request was 
turned down. Hence, for up-to-date data 
on the activities and progress of the Japa- 
nese garrison, reliance had to be placed en- 
tirely upon photographic reconnaissance, 
chiefly aerial. 65 

Not until 25 April, after Conolly's staff 
had arrived on Guadalcanal, were photo- 
graphs received, and the first ones were 
badly obscured by cloud cover. Later, 
aerial photographs were only fair, but were 
supplemented by excellent obliques of the 
coast line taken by the submarine USS 

" :l Takeda, Outline of Japanese Defense Plan. 
K4 MIS WD, Survey of Guam ONI 99. 
65 III Phib Corps Rpt Guam, Intel, p. 3. 



Greenling. Maps of the interior, prepared 
from prewar sources and revised on the 
basis of these aerial photographs, were 
fairly good as to scale and azimuth, but 
only occasionally did they portray ground 
contours accurately. Changes made by 
the Japanese in the road system were not 
indicated on the maps provided the troops; 
in fact, in the north of the island map 
locations of the roads were as much as 
i ,500 yards off from their true positions. 
Trails were not shown at all. 66 

With the capture of Saipan, a good 
number of Japanese documents were made 
available to the planners for Guam and 
afforded them for the first time some idea 
of the enemy situation. On the basis of 
these documents and of interrogations of 
prisoners of war, the intelligence section 
of III Amphibious Corps estimated that 
Guam was garrisoned by a total of 17,673 
Army troops and 9,945 to 10,945 Navy, 
Air, and construction personnel. Although 
these figures proved to be considerably in 
excess of actual enemy strength, Gcigcr's 
staff correctly predicted that the bulk of 
the Army troops on the island was com- 
posed of the sgth Division under Lt. Gen. 
Takeshi Takashima and the nth Infantry 
Regiment under Maj. Gen. Kiyoshi Shigc- 
matsu." 7 The principal naval unit on the 

KK Ibid., p. 1; TF 53 Opn Rpt Guam, 10 Aug 
44, Incl E, p. a. 

fi7 The nth Infantry Regiment, somewhat ex- 
panded, had actually been renamed the 4.8th In- 
dependent Mixed Brigade. 

island was thought to be the 54th Naval 
Guard Force, and it was believed that 
about 2,185 naVa l air personnel were sta- 
tioned there as well. 68 

It was assumed, on the basis of these 
documents and interrogations, that the Jap- 
anese would concentrate their defenses 
around Tumon Bay, Agana, and Agat, all 
on the west coast. Only two battalions 
were thought to be garrisoned on the south 
and southeast coasts. American troops 
about to invade Guam were warned to ex- 
pect a large amount of mobile artillery and 
a determination on the part of the Japa- 
nese to exploit the mountainous terrain, 
which provided excellent observation. The 
enemy was thought to be holding back 
(from the beach defense) a mobile reserve 
of at least one battalion plus supporting 
weapons in the Agana area. A smaller re- 
serve of about reinforced company strength 
was believed to be located somewhere in- 
land of Agat. 

"It seems evident," concluded Geigcr's 
intelligence section, "that both we and the 
Japanese have been thinking along the 
same lines, that is, the beaches we find best 
for landings are those the Japs find most 
dangerous to them and have fortified the 
most." e " The conclusion was fully war- 

08 III Phib Corps Opn Plan 1-44, Supple- 
ment 4 to Annex B, 9 Jul 44. 
69 Ibid., p. 5. 


The Enemy 

The Japanese defense of Guam was 
much less effective than that of Saipan. 
Not only did the Japanese have fewer men, 
less artillery, and fewer tanks than their 
compatriots on LSaipan but they also had 
a much larger area of land to defend. 
Nevertheless, they had ideal terrain for the 
defense and a sufficient force to prevent a 
rapid or easy conquest of the island. 

Guam's defense was commanded initially 
by General Takashima, Commanding Gen- 
eral, sgth Division and Southern Marianas 
Army Group. In the middle of June Gen- 
eral Obata, Commanding General, 31st 
Army, reached Guam with his two senior 
staff officers. He had been in the Palaus, 
probably because the Japanese expected 
the next American thrust to be in that 
area. Once it became apparent that the 
blow would come farther north, he had 
hastened to the Marianas, too late, how- 
ever, to reach his headquarters on Saipan. 
Instead, he landed on Guam to linger in 
forced inactivity while the garrison on Sai- 
pan went down to defeat. His presence on 
Guam had very little influence on Japanese 
tactics there until the death of General Ta- 
kashima on 28 July, after which Obata as- 
sumed direct command. 1 

Troops and Troop Dispositions 

In mid-July, on the eve of the American 
invasion of Guam, the Japanese defenders 
numbered about 18,500 men. The early 
preponderance in air and naval strength 
that the Americans were able to establish 
in the area resulted in the loss to the Guam 
garrison of about 900 much needed men, 
including the 1st Battalion, 10th Independ- 
ent Brigade. These troops had been tem- 
porarily stationed on Rota and on 8 June 
had been ordered to return to their parent 
commands on Guam. By the time the move 
could be organized, however, the 38-mile 
stretch of water between the two islands 
was under close American surveillance, 
and the transfer was never made. 

All together, the American invaders 
faced an understrength garrison composed 
of eleven Army infantry battalions, two 
and two-thirds Army artillery battalions, 
three tank companies, two Army antiair- 
craft companies, Army engineers, service 
troops, and so forth, together with various 
Navy units, the most important of which 
were the 34th Naval Guard Force and the 
60th Antiaircraft Defense Unit? 

In early June these forces were spread 
all over the island as a precaution against 

1 Japanese Studies in World War II, 5f> : 
56 Rpt Forager, Incl D, G-2 Rpt. 

TF a Sec Appendix F for complete Japanese order 

of battle. 


E*5 322 

f.Mt Barrigotfa 

A PR A /^ m 3 HS) 3 ?- lli Smajana rf 

-^ HARBOR ) IS 181 -' .-^CS3 i0i -o 9 r,±,\» 

Santo Rota ' 

8 /?? 


Asinaft L, 



MffO MC 


July 1944 

NGF N»v»L Guard fosce 
£tevetiens m teet 


AM/ 5 17 



an invasion from any direction. Guam was 
divided into four sectors for purposes of 
defense. In the Agana sector were stationed 
the four battalions of the 48th Independ- 
ent Mixed Brigade and the 3d Battalion, 
38th Infantry; in the area around Agat 
were the 1st and 2d Battalions, 38th In- 
fantry; on the south coast in the Inarajan 
sector were two battalions of the 10th In- 
dependent Mixed Brigade; and in the 
northern sector, with headquarters at 
Finaguayac, was the 2d Battalion, 18th 

By July this picture had radically 
changed. The American naval shelling of 
Agat on 16 June had tipped off the Jap- 
anese as to the probable place of the forth- 
coming landings, and the postponement of 
W Day gave them ample opportunity to 
reorganize their defenses. By mid- July al- 
most the entire garrison had been moved 
to the w est coast be tween Agat and Tu- 
mon Bay. \(Map ij)\ 

At the time ol the American amphibious 
assault, Headquarters, 2gth Division, and 
most of the division's service troops were 
located at Fonte, as was the headquarters 
of the 48th Independent Mixed Brigade. 
General Shigcmatsu, Commanding Gen- 
eral, 48th Independent Mixed Brigade, 
commanded the Agana sector, which 
stretched along the west shore from Piti 
to Tumon Bay and included the great ma- 
jority of the Japanese forces on the island. 
For purposes of shore defense, the Agana 
sector was divided into three, perhaps four, 
beach defense areas. From northeast to 
southwest, the first of these was at Tumon 
Bay, where the 323d Independent Infantry 
Battalion was located. The 321st Independ^ 

:l TF 56 Rpt Forager, Incl D, G-2 Rpt, p. 47. 
The locations of the gd Battalion, 18th Infantry, 
and of the several naval units were unknown, 

ent Infantry Battalion defended the area 
around Agana Bay, and the 320th Inde- 
pendent Infantry Battalion manned the 
defenses between Adelup Point and Asan 
Point, where the 3d Marine Division was 
to land. In the Piti area was the 18th In- 
fantry Regiment, less the 1st Battalion, 
which was on Saipan. The unit was con- 
siderably understrength since some of its 
personnel and much of its equipment had 
been lost when one of its ships en route 
from Japan had been sunk by an Ameri- 
can submarine. The 18th Regiment also 
had partial responsibility for Asan Point in 
case the Americans should land there. 

The 10th Independent Mixed Regiment, 
less its 1st Battalion and gth Company, 
was in the Fonte-Ordot area. The 3i$th 
Independent Infantry Battalion was inland, 
cast of Agana, in reserve. Two of the three 
tank units on the island were also in re- 
serve, poised to strike the beachhead with 
the infantry. These were the 29th Division 
Tank Unit at Ordot and the 2d Company, 
Qth Tank Regiment, at Sinajana. Also in 
general reserve was the Otori Unit, com- 
posed chiefly of naval air personnel reor- 
ganized into a jerry-built unit for ground 
combat. Most of the Army artillery, in- 
cluding the 48th Independent Mixed Bri- 
gade Artillery Unit, formerly the 3d Bat- 
talion, nth Mountain Artillery Regiment, 
was disposed throughout the Agana sector. 
The two batteries of the ioih Independent 
Mixed Regiment Artillery Unit had been 
removed from regimental control and 
placed directly under the 48th Independ- 
ent Mixed Brigade. Certain guns of these 
batteries were located just inland of Tu- 
mon Bay, but the majority were in the 
vicinity of Agana, The 38th Infantry's Ar- 
tillery Battalion was broken up, one bat- 
tery attached to each infantry battalion, so 



that the 3d Battery, attached to the 9J 
Battalion, was also in the Agana sector 

The Agat sector was commanded by Col. 
Tsunetaro Suenaga, commanding officer of 
the 38th Infantry Regiment, whose com- 
mand post was on Mount Alifan. The 
1st Battalion, 38th Infantry, covered the 
beaches in the Agat area, and the 2d Bat- 
talion of the same regiment occupied the 
base of Orote Peninsula. To the rear of the 
Agat beaches, the 1st Company, gth Tank 
Regiment, was in readiness to counterat- 
tack in case of a landing. Also in reserve 
for the Agat beaches was the gth Com- 
pany, 10th Independent Mixed Regiment. 
Orote Peninsula was garrisoned by the 
main body of the 54th Naval Guard Force; 
the 755th Air Unit, reorganized for ground 
combat; and the two batteries of the 52 d 
Antiaircraft Artillery Battalion, which was 
charged with antiaircraft defense of Orote 
airfield. Since this unit's guns, which were 
75-mm. antiaircraft, could be depressed as 
low as minus seven degrees, they were in 
eflect dual purpose and could be used to 
supplement the conventional field artillery 
and antiboat weapons. 4 

Other units, such as service, engineer, 
and construction units, were scattered 
throughout the island, some on the west 
coast, and some inland as far as Santa 
Rosa. None of these, however, had any 

A Sources from which Ihe locations of the Jap- 
anese units on Guam were derived are as follows: 
TF 53 Rpt Forager, Incl D, C-2 Rpt; Japanese 
Studies in World War II, 55; Ltr, Takcda to Mc- 
Queen, USMC; III Phi b Corps Rpt Guam, Incl 
D, G-2 Rpt; CINCPAC-CINCPOA Trans 10194, 
Opt! Orders of Guam Garrison Units, 15— 16 Jul 
44; 77th Inf Div G-2 Jul File, Guam Opn; POW 
Interrogation Rpts made by III Phib Corps, Ilq 
3d Marine Div, SLF (TG 56.2), and Intel Sec 
2rst Marines, in Records Sec Hist Branch G-3, 

significant combat value. Nor could the 
Japanese arm the civilian population, most 
of which appears to have remained at least 
passively loyal to the United States. As of 
10 January 1944, the native Guamanians 
numbered about 24,000. 5 Slightly over a 
hundred were of mixed American and 
Chamorro parentage and had been jailed 
as soon as the Japanese occupied the is- 
land. The rest of the population suffered 
some organized maltreatment and abuse in 
the early days of Japanese rule, but this 
appears to have gradually tapered off. 
However, rigid food rationing, forced labor, 
confiscation of property without compensa- 
tion, exclusion from business enterprises, 
and a score of lesser deprivations and 
humiliations kept the native population 
sullen and restive during the period of Jap- 
anese occupation. In June 1943 all able- 
bodied men between the ages of fourteen 
and sixty were forced to work for the oc- 
cupation army, and women were ordered 
to replace the men in the fields. After the 
American air raid of 1 1 June, large num- 
bers of natives fled to the hills. Many were 
rounded up by Japanese military police 
and placed in camps near Asinan, Manen- 
gon, and Talofofo. The Guamanians were 
clearly poor raw material for collaboration- 
ism, and there is no evidence that the 
Japanese made any successful attempt to 
reconstruct them to that end. 6 

Japanese military doctrine for the de- 
fense of Guam was essentially the same as 
that for Saipan. Emphasis was placed on 
meeting and annihilating the enemy at the 

B III Phih Corps Rpt Guam, Incl D, G-2 Rpt 

" Sources for the description of the lot of Guam's 
civilians under Japanese rule are William Hip- 
pie's report in Newsweek, August 21, 1944, p- 35, 
and Thompson, Guam and Its People, p. 160. 



Table 2 — Artillery on Guam 

Type of Weapon 

Estimate of 

III Amphibious 

Corps (t-2 


Estimate of 
Colonel Takcda 

20-cm. coastal defense guns (short Navy guns) 

15-cm. coastal defense guns 

12.7-cm. coastal defense guns 

12-cm. coastal defense guns 

12-cm. dual-purpose guns 

105-mm. howitzers (M91) 

75-mm. mountain and field guns 

75-mrn. antiaircraft, guns 

57-mm. antitank guns 

37-mm. antitank guns 

25-mm. antiaircraft guils (Navy) 

20-mm. antiaircraft guns (Army) 











Source: III Pliiti Corps Rpt Guam, Incl D, G-2 Rpt; Ltr, Takcda to McQueen, 4 Oct 46- The first column represents t.Iie Japanese pieces 
captured or destroyed before 10 August, the date Guam was officially declared secured. ColonelTakeda's estimate was made in October 1946, 
more than two years after the battle. Of the two eHtiniates, the first is undoubtedly the more accurate. 

beaches. If that failed, an organized coun- 
terattack was to be delivered against the 
beachhead soon after the landing. Finally, 
if the invaders succeeded in establishing 
and holding their beachhead line, the Jap- 
anese would retire to the hills and fight on 
from there. 

Thus General Shigematsu, in command 
of the Agana sector garrison unit, which 
contained the majority of troops on the is- 
land, declared on 15 July, "It has been 
decided that the enemy is going to launch 
an attack in force at dawn in the region 
of the Agana sector. When he lands, the 
Division will be quick to seize the oppor- 
tunity to attack him in this sector with a 
powerful force and crush him at the 
beaches. . . . The Garrison Unit will await 
its initial opportunity and will completely 
destroy the enemy landing force upon 
the beaches." 7 If the infantry units at 


the shore line failed in their mission, the 
10th Independent Mixed Regiment, two 
battalions of the iSth Infantry, and the 
3d Battalion, gQth Infantry, were ordered 
to carry out the second phase of the plan 
— a counterattack in force against the 
American beachhead in this area. 

Supporting Weapons 

Artillery on Guam, including coast de- 
fense, field artillery, and antiaircraft and 
antitank weapons, was manned by both 
Army and Navy personnel. As indicated 
above, the bulk of the field artillery pieces 
in July were situated to command the 
western shore line from Tumon Bay to 
Agat. Table 2 gives two estimates of the 
number and type of these weapons. Just 
how many were still operational at the time 
of the American landings is not known, and 
even a comparison of these figures with 
the assessment of damage wrought by the 



preliminary bombardment cited above will 
provide only a hazy idea of Japanese ar- 
tillery strength on 2 1 July. 

Japanese tank strength was considerably 
lower than it was believed to be by Amer- 
ican intelligence staffs, both before and 
after the battle. Although the Americans 
claim to have destroyed or captured fifty- 
nine enemy tanks by 1 1 August, actually 
there were no more than thirty-eight pres- 
ent at any time, and possibly even fewer. 8 
The ist Company, gth Tank Regiment, 
which was located in the Agat-Orote area, 
had from twelve to fifteen light tanks. The 
sd Company of the same regiment and the 
2yth Division Tank Unit, both situated so 
as to support the defense of the Asan 
beaches, had a total of from twenty-one to 
twenty-three tanks, of which at least ten 
were mediums. 


The main fortified area ran along the 
west coast from Tunion Bay to Facpi 
Point and included, of course, Orote Pen- 
insula. Other fortified beaches, on the 
south and east coasts from Merizo to Pago 
Bay, had been abandoned before W Day, 
their defenders having moved to the north. 
Outside the main fortified area, the air- 
fields were provided local defense by anti- 
aircraft and dual-purpose guns. 

The most notable and certainly the 
most effective fortifications on the island 
were constructed across the neck of Orote 
Peninsula, which contained a fairly elabo- 
rate system of trenches and foxholes 

8 TF 56 Rpt Forager, Incl D, G-a Rpt, p. 
65; 3d Marine Div Interrogation Rpt 396, Capt 
Sato, Hideo (CO 34th Tk Co, 29th Div Tk Unit), 
Records Sec Hist Br G-3, Hq USMC; CINCPAG- 
CINCPOA Trans 9304, 9th Tk Rgt Order of 

arranged in depth, together with large 
numbers of pillboxes and heavy-caliber 
weapons. Outside of Orote, the prepared 
defenses were generally hastily constructed 
and often incomplete. The typical beach 
defense was arranged, from the seaward 
side, in four parallel lines: first were ob- 
stacles and mines on the fringing reef off- 
shore; second came beach obstacles and 
tank traps; third were trenches, machine 
gun positions, pillboxes, heavy weapons, ar- 
tillery, and coast defense guns on the 
beaches or immediately inland; and, finally, 
came the machine guns, heavy weapons, 
and artillery emplaced on the high ground 

Insufficient advantage was taken of the 
high ground, and except on Orote little 
provision was made for defense in depth. 
Even as late as the five-week period of pre- 
invasion bombardment, the Japanese con- 
tinued to work frantically on improving 
offshore obstacles and beach defenses, to 
the neglect of positions in the rear. 9 

During the first two years of the war, 
the Japanese had slighted the military de- 
velopment of the Marianas in favor of more 
forward areas, and almost nothing was 
done to fortify Guam until early 1944. The 
beaches had to be protected first, and 
Guam's large size and numerous possible 
landing points meant that proportionately 
greater effort had to be expended at the 
shore line before work could be com- 
menced on defenses in depth. The time, 
effort, and materials expended on the south 
and southeast shores was ultimately wasted 
when these positions were abandoned. Af- 
ter the assault on Saipan, Guam was en- 
tirely cut off from its sources of materiel, 

! ' Ltr, Takeda to McQueen, 4 Oct 46. 



and the five-week period of bombardment 
not only destroyed many of the existing 
fortifications but also severely hampered 
efforts on the part of the Japanese to con- 
tinue construction. 10 

As has been already mentioned, the 
Japanese relied heavily on coral-filled palm 
cribs and wire cages to interrupt and im- 
pede the approach of landing craft to and 
over the offshore reef. 11 In addition to 
these, which were all blown up by Amer- 
ican underwater demolition teams, a series 
of antiboat mines of about forty to fifty 
pounds was placed along the reef or be- 
tween the obstacles. The beaches them- 
selves were strung with barbed wire and 
in some places aerial bombs were embedded 
in the sand just inland of the wire. Anti- 
tank obstacles also were installed on the 
beaches by lashing coconut logs across 
trees or setting them vertically in the 

The Agat beach defenses were typical of 
the others. Here was an almost continuous 
line of open trenches about two feet wide 
and three and a half feet deep. Running 
parallel to the shore line approximately 
fifty feet inland of the high-water mark, 
these trenches were supplemented by an 
occasional rifle or machine gun pit about 
eight feet to their front and by shelters to 
their rear. Distributed along the beach be- 
tween Agat and Bangi Point were about 
twenty-five pillboxes. A strong concrete 
blockhouse on Gaan Point held two 75- 
mm. mountain guns, one 37-mm. gun, and 
positions for machine guns and riflemen. 
Two concrete emplacements of 40-mm. 
guns were located between Gaan Point and 

10 TF 56 Rpt Foragkr, Incl D, G-a Rpt, 

Agat only about five feet inland of the 
high-water mark. 12 

Pillboxes here were constructed of palm 
logs, sandbags, reinforced concrete, earth, 
and coral rocks, the majority being simple 
structures of palm logs covered with earth. 
They averaged about eight feet square and 
three feet high, with roofs two feet thick 
in the center and one foot at the edge. 
Usually, they had two firing ports about 
twelve by four inches in size, which allowed 
only a fairly narrow traverse. The rein- 
forced concrete types were either square or 
octagonal and located chiefly along the 
roads. About eight feet across and two to 
three feet high, their walls were about six 
inches thick. There were a few masonry 
pillboxes of coral rocks with walls a foot 
thick. All the pillboxes on the beaches 
were mutually supporting against attack 
from the seaward side, but not against at- 
tack from the flanks or rear. 

Japanese Situation on the Eve of Battle 

The effects of American preliminary 
bombardment on Japanese defenses has al- 
ready been described in as much detail as 
surviving records permit. The devastation 
was great and widespread, if not as effec- 
tive as the invaders at first believed. More- 
over, the accelerating tempo of naval shell- 
ing and aerial bombing and strafing made 
it almost impossible for the Japanese to re- 
pair the damage or to engage in new 

Notwithstanding these difficulties, the 
defenders bent every effort to shore up 
their crumbling defenses to the last minute 
before the invasion. But most of the labor, 

pp. 61-63. 

11 See above, p. 322 

12 MIS WD Mil Rpt 25, Jan 45, p. 31. 



Japanese Heach Obstacles at Guam were numerous along the northern landing 

at least along the beaches, had to be per- 
formed at night when darkness and the 
physical exhaustion of the troops slowed 
progress to a snail's pace. The dilemma 
was inescapable, as is attested by one Jap- 
anese Army lieutenant who complained, 
"Our positions have been almost com- 
pleted but they have not been done as we 
had hoped . . . great effort was put into 
the construction but we still have been un- 
able to complete the cover. We are in a 
terrible fix." 13 

The weeks of American bombardment, 
the prolonged uncertainty, the anxious 
waiting from day to day while explosions 
rent the air but no American soldier came 
into view to be shot at or stabbed with 
bayonet- — -all these factors took their psy- 

chological toll even on the martial-minded 
Japanese. They suffered greatly in seishin 
— a word that means not so much "mor- 
ale" as "psychological well-being." After 
several days of successive attacks, "scat- 
tered outbreaks of serious loss of spirit" 
occurred. After another week the spirit of 
some of the men deteriorated so badly that 
they "could not perform their duties in a 
positive manner." 14 

This cumulative physical and psycholog- 
ical exhaustion would show up in the days 
of battle to come. The troops on Guam 
tended to become more easily disorganized 
than had their compatriots on Saipan and 
Tinian. They turned more readily from or- 
ganized combat to futile and suicidal in- 
dividual displays of fanaticism. 13 

13 Imanishi Diary; see also C I N C P A C- 
CINCPOA Trans 10802, extracts from the diary 
of Cpl Suzuki, Tai; GINCPAC-CINCPOA Trans 
1 04 10, diary of an unidentified soldier. 

11 The quotations are from Ltr, Takeda to Mc- 
Queen, 20 Feb 52. 

1S III Phib Corps Rpt Guam, Incl D, G-2 
Rpt 7- 



Japanese Open-Trench Beach Defenses at Agat, where the 1st Provisional 
Brigade units landed. 

But if their seishin was ebbing, the Jap- 
anese on Guam remained high in shiki— 
meaning morale in the sense of a willing- 
ness to die in combat, 16 This spirit is 
reflected, with the usual rhetorical flour- 
ishes, in the diary of one enlisted man: "I 
will not lose my courage, but now is the 
time to prepare to die! If one desires to 
live, hope for death. Be prepared to die! 

With this conviction one can never lose 
.... Look upon us! We have shortened 
our expectancy of 70 years of life to 25 in 
order to fight. What an honor it is to be 
born in this day and age." 1T 

Against this kind of determination, the 
task facing the marines and soldiers of the 
III Amphibious Corps would by no means 
be light. 

lu Ltr, Takeda to McQueen, 24 Jan 52. 

17 CINCPAC-CINCPOA Trans 10802. 

V Holm.., Jr 

MAP 18 


The Fight for the Beachhead 

As evening of 20 July closed in, Admiral 
Conolly and those around him who were re- 
sponsible for the invasion of Guam looked 
forward with high optimism to the success 
of their enterprise. Reports indicated that 
"all known major defensive installations 
in position to interfere with the transports 
approach and the landing have been de- 
stroyed and the assault beaches cleared of 
obstructions and searched for mines with 
negative results." 1 The weather forecast 
for 21 July was excellent. Conolly con- 
firmed W Day as 21 July and H Hour as 
0830. To his entire task force he prom- 
ised, "conditions are most favorable for a 
successful landing." 

Ashore, the Japanese command was not 
to be outmatched, at least in outward show 
of confidence. To the men of the 48th In- 
dependent Mixed Brigade, General Shige- 
matsu announced encouragingly: "The 
enemy, overconfident because of his suc- 
cessful landing on Saipan, is planning a 
reckless and insufficiently prepared landing 
on Guam. We have an excellent opportun- 
ity to annihilate him upon the beaches. We 
are dedicated to the task of destroying this 
enemy, and are confident that we shall 
comply with the Imperial wish." 2 

W-Day Preliminary Bombardment 

The morning of 21 July dawned bright 
and clear as predicted. The wind was light, 
the sea calm. At 0530 came the first order 
to commence fire and for almost three 
hours the din and smoke of incessant naval 
salvos filled the air over the transport areas 
and shot up great geysers of dirt and debris 
ashore. From their flagships Appalachian 
and George Clymer, Admirals Conolly and 
Reifsnider directed the bombardment of 
the Asan a nd Agat beaches, respectively. 
(M ap 18 J Ships' fire was slow and delib- 
erate, concentrating on the landing beaches 
themselves, their flanks, and the areas im- 
mediately inland. On this day alone the 
fire support ships would expend a total of 
342 rounds of 16-inch shells, 1,152 rounds 
of 14-inch, 1,332 rounds of 8-inch, 2,430 
rounds of 6-inch, 13,130 rounds of 5-inch, 
and 9,000 4.5-inch rockets. 3 

Twenty minutes after the opening of this 
tremendous naval bombardment, the first 
planes appeared overhead to provide cover 
against possible enemy air or submarine at- 
tack. At 0615 a roving patrol of twelve 
fighters, nine bombers, and five torpedo 
planes from the carrier Wasp made the 
day's first strike — against tiny Cabras Is- 
land lying just off the right flank of the 3d 

1 TF 53 Rpt Guam, Irtcl A, Opns Chronology, 
p. 11. 

3 TF 56 Rpt Foragkr, Incl G, Naval Gun 
2 III Phib Corps Rpt Guam, Incl D, G^2 Rpt 1. Fire Rpt, p. 71. 



First Wave of Landing Craft Heads for Agat Beaches. Smoke is from pre- 
invasion naval shelling and air bombardment. 

Marine Division's zone of attack. These 
and all other planes from Mitscher's Task 
Force 58, as well as those flown from the 
escort carriers of Task Force 53, were un- 
der the control of Conolly's Commander, 
Support Aircraft, aboard Appalachian. 
Also on station was an airborne air co- 
ordinator to direct scheduled air strikes 
and such other attacks as might later be 
ordered by the Commander, Support Air- 
craft. 4 

4 TF 53 Rpt Guam, Incl A, p. 26; Ibid., Incl 
C, p. 13. 

From 0715 to 0815 planes from Mits- 
cher's fleet bombed and strafed the four- 
teen miles of coast line from Agana to 
Bangi Point. During this period of simul- 
taneous attack from air and sea, naval guns 
were restricted to a maximum range of 
8,000 yards to insure that the ordinate of 
their shells would be no higher than 1,200 
feet. Pilots were instructed to pull out of 
their runs before reaching as low as 1,500 
feet — a precaution taken to save Ameri- 
can planes from being struck by friendly 
naval fire while at the same time rendering 
unnecessary the usual cessation of naval 



Asan's Green and Blue Beaches stretch from Asan town (left center) to Asan 
Point (lower right). 

fire during the aerial strikes. Throughout 
this hour of bombardment, 312 of Mits- 
cher's planes pounded the northern and 
southern landing beaehes and their flanks 
with a total of 1 24 tons of bombs/' 

Over each of the two sets of beaches 
flew a naval observation plane equipped 
with parachute flares, which were released 
when the troop-laden landing vehicles were 
1,200 yards (8 minutes) from the shore 
— a signal to the support ships to step up 
their fire. Immediately, the support ships 

1 TV 53 Rpt Guam, Incl C, p. 13. 

commenced a last mighty bombardment of 
the beaches and maintained it until the 
troops were reported to be only 300 yards 
from the shore. Then all naval fire was 
shifted inland, to continue even after the 
first of the marines had touched land. The 
parachute flares were also the signal for a 
special H Hour strike by aircraft from 
Task Force 53. A flight of forty-four 
fighters strafed the two landing beaches 
until the first troops were almost ashore, 
then shifted their attack 1,500 to 2,500 
yards inland. This strike was followed by a 
flight of twenty-four, whose task was to 



keep the enemy down until the first wave 
of marines could cross the open beaches 
and gain cover. 6 

From Ship to Shore 

In the midst of all this fire and smoke, 
the ships carrying the assault troops arrived 
on schedule and took station in the trans- 
port areas off the landing beaches. Am- 
phibian tanks and amphibian tractors 
laden with marines eased their way into 
the water through the open bow doors 
of the LST's, transports lowered their 
LCVP's over the sides, and the LSD's let 
down their tail gates to permit their tank- 
carrying LCM's to become waterborne. 

The pattern of the ship-to-shore move- 
ment was essentially the same for the 
northern and southern beaches — a pattern, 
in fact, that had been standardized in the 
Pacific since the invasion of the Marshalls. 
Ahead of the first waves of landing vehicles 
bound for each set of beaches went a line 
of nine LCl(G)'s firing 4.5-inch rockets 
and 40-mm, and 20-mm. guns as they came 
within range of the shore line. Then came 
a wave of amphibian tanks firing 37-mm. 
and 75-mm. guns to escort the marines 
over the reef and onto the beaches. Close 
astern were the successive waves of 
amphibian tractors carrying the assault 
troops, followed at intervals by LCM's em- 
barking tanks. Reserve troops were loaded 
in LCVP's that would have to stand by at 
the edge of the reef until the first waves 
of tractors had debarked their troops and 
could return to take more men in from the 
reef. 7 

6 Ibid., p. 14. 

1 The formation of the landing waves is given 
in SLF TG 56.2 Opn Plan 1-44, 1 1 May 44, 
Annex G, Charts n and 12. 

Off the northern beaches the LCI(G)'s 
crossed the line of departure at 0806 and 
commenced firing their rockets when they 
were about 1,000 yards off the shore line. 
At 0819 the first wave of troop-carrying 
LVT's reached a line approximately 
1,200 yards offshore. Parachute flares were 
dropped from the observation plane over- 
head, and three minutes later the last 
flurry of intense close naval fire on the 
beaches began. The tempo of firing of the 
5-inch guns increased from five to ten 
rounds per minute. About 300 yards from 
the shore the LCI(G)'s ceased their salvos 
of rockets, moved to the flanks, and, along 
with the destroyers, kept these areas under 
fire. By then mortar and artillery fire from 
shore was falling among the approaching 
tractors. In spite of the pounding they had 
been receiving from ships and aircraft, the 
Japanese were able to bring their pieces 
to bear on the assault waves and destroy 
nine of the 3d Marine Division's amphibian 
tractors. 8 Nevertheless, the waves of ve- 
hicles rolled on, and at 0828 the first 
LVT(A) touched down, two minutes 
ahead of schedule, 

Off Agat, fire from the shore was even 
heavier. There, the Japanese greeted the 
invaders with a hail of small arms, ma- 
chine gun, and mortar fire for the duration 
of their approach to the shore. In addition, 
several 70-mm. guns in well-placed con- 
crete blockhouses enfiladed the beaches and 
fired at the LVT's as they crossed the reef. 
On Gaan Point, near the middle of the 
landing area, a 75-mm. field piece and a 
37-mm. gun opened fire as did a 75-mm, 
field piece on Yona Island to the right. 
One LCI(G) was hit and thirteen am- 

H 3d Marine Div War Diary, Jul 44, p. 6. 
9 TF 53 Rpt Guam, Incl A, pp. 27-30. 



3D Marine Division Beachhead. All types of invasion craft in background; 
first-aid station in foreground. 

phibian tractors were knocked out. 10 Still 
the landing formation held, and the troops 
got ashore on schedule. 11 In the zone of 
the 4th Marine Regiment (and there alone 
along the entire corps landing zone) the 
amphibian tractors had been ordered to 
move a thousand yards inland before dis- 
embarking their troops. As on Saipan, this 
scheme failed, and here as elsewhere during 
the assault on Guam, the leading waves of 

10 TG 53.2 Rpt Guam, Incl D, p. q. 

11 1 st Provisional Marine Brig Opn Rpt and 
SAR Jnl, p. 10. 

troops dismounted from their LVT's close 
to the shore line. 12 

The Northern Beaches 

The 3d Marine Division plan called for 
a landing of the assault elements of all 
three regiments over about 2,500 yards of 
beach between Asan Point and Adclup 
Point. From right to left, the 9th Marines 
was to go ashore on Blue Beach, the 21st 
Marines on Green Beach, and the 3d Ma- 

1 3 1 st Provisional Marine Brig War Diary, 
Jul-10 Aug 44, p. 4. 



rines on Red Beaches i and 2. On hitting 
shore, the division was to capture the cliffs 
and high ground immediately inland and 
prepare for further operations to the east 
and southeast. The 9th Marines was to 
seize the low ridge facing its beach, protect 
the division right flank, and be prepared to 
move along the coast to take Piti Navy 
Yard and Cabras Island if so ordered. In 
the center, the 21st Marines was to capture 
the cliff line to its front. On the left the 
3d Marines was ordered to secure Adelup 
Point, Chonito Cliff, a nd the high ground 
southeast, of the cliff. l3 \(Map VI. )\ 

On the right, the 9th Marines, com- 
manded by Col. Edward A. Craig, moved 
ahead steadily against fairly heavy machine 
gun and rifle fire and by the end of the 
day had secured a beachhead about 1,500 
yards in depth. 14 In the center, Col. Ar- 
thur H. Butler's 21st Marines found the 
opposition only moderate, but the terrain 
"almost impossible." Ahead lay low cliffs 
of bare rock mounting to a plateau covered 
with scrub growth and tangled vines. In- 
tense heat and a shortage of water added 
to the day's ordeal, but in spite of the 
natural obstacles as well as enemy mortar 
and rifle fire, two battalions had made their 
way to the top of the cliff and tied in 
with the 9th Marines on the right by night- 

It was on the division left that the day's 
most intense and arduous fighting oc- 
curred. Northeast of Red Beach 1 was 
Chonito Cliff, containing a complex system 
of cave defenses and offering the enemy a 

IS 3d Marine Div Opn Plan 2-44, 13 May 44. 

14 This account of 3d Marine Division's opera- 
tions is derived from: 3d Marine Div SAR For- 
ager Opn, 19 Aug 44, p. 1 ; Ibid., Incl L, 3d 
Marines SAR, p. 1 ; Ibid., Incl M, 9th Marines 
SAR, p. 1; Ibid,, Incl M, 21st Marines SAR, p. 
1 ; Lodge, Guam, pp. 38-47. 

perfect opportunity to pour mortar and ar- 
tillery fire on the low ground below. Inland 
of Red Beach 2 lay Bundschu Ridge, ls of- 
fering the same sort of obstacles. By noon, 
Chonito Cliff was cleared, but two separate 
and costly attacks on the ridge to the south 
failed to make any significant headway, 
and by nightfall the right flank of Col, W. 
Carvel Hall's 3d Marines was still out of 
contact with the 21st Marines in the di- 
vision center. 

The day's casualties on the northern 
beachhead came to 105 killed, 536 
wounded, and 56 missing in action. The 
cost was more than had been anticipated, 
yet the results, if far from decisive, were 
at least hopeful. The beaches themselves 
were secured, as was Chonito Cliff on 
the left flank. By late afternoon Gen- 
eral Turnage, the division commander, was 
ashore, and his command post was in op- 
eration; division artillery was in position 
and in the process of registering; division 
engineers had cut a road to supply the 
2 1 st Marines and were at work removing 
land mines and demolishing cave positions. 

The Southern Beaches 

According to the operation plan of the 
1 st Provisional Marine Brigade, the two 
Marine regimental combat teams would 
land in assault, and Col. Vincent J. Tanz- 
ola's 305th Regimental Combat Team of 
the 77th Infantry Division would be in re- 
serve. The 4th Marines on the right was 
to go over White Beaches i and 2 north 
of Bangi Point, establish a beachhead, and 
protect the brigade right flank. On the left, 
the 2 2d Marines was to land on Yellow 

18 So named after Capt. Geary R. Bundschu, 
Commander, Company A, 3d Marines, who later 
lost his life there. 



Beaches i and 2, occupy the town of Agat, 
and then drive north to seal off Orote Pen- 
insula at its base. The 305th Infantry, 
when committed to action on brigade or- 
der, would pass through the lines of the 
4th Marines and p rotect the r ight sector 
of the beachhead. 1 6 | (Map VII. j| 

The terrain facing the brigade was far 
more favorable for the attack than that 
facing the 3d Marine Division to the north. 
Low hills and open ground characterized 
most of the area immediately inland of the 
landing beaches. In spite of numerous pill- 
boxes still alive with Japanese, both of the 
assault regiments moved forward rapidly. 
By early afternoon, the 4th Marines had 
cleared Bangi Point and Hill 40 just inland 
and had set up a roadblock supported by 
five tanks on Harmon Road, which ran 
east out of Agat. The 22d Marines cap- 
tured Agat, which had been mostly re- 
duced to rubble by naval and air bom- 
bardment but still contained its share of 
defenders hiding in the debris. Moving out 
from Agat, the regiment began to receive 
artillery fire from the hills beyond and suf- 
fered some casualties, as well, from a mis- 
placed American air attack. It was also 
handicapped by the fact that a communi- 
cations failure kept the reserve battalion 
(the 3d) afloat until too late to participate 
in the day's fighting. 17 

Despite these difficulties, by the close of 
the first day's fighting the brigade had es- 
tablished a beachhead ranging in depth ap- 
proximately 1,300 to 2,300 yards at the cost 
of only about 350 casualties. Brig, Gen, 
Lemuel C. Shepherd, Jr., came ashore be- 

fore noon and his command post southeast 
of Gaan Point was in operation by 1350. 
Defenses were organized in depth against 
the expected Japanese night counterattack. 
Artillery was ashore and registered. 

Landing the 305th Infantry 

The operation plan for the 1st Provi- 
sional Marine Brigade had designated the 
305th Regimental Combat Team as bri- 
gade reserve, to be used at General Shep- 
herd's discretion. Accordingly, the brigade 
commander had ordered one battalion 
team (later designated the 2d, commanded 
by Lt. Col. Robert D. Adair) to be boated 
and ready at the line of departure by 1030 
of the 2 i st. The other two battalion land- 
ing teams would follow the 2d ashore when 
and as needed, 18 

Actually, the first element of the 77th 
Division to go ashore was a liaison party 
of four from the 242d Engineer Combat 
Battalion, which was attached to the 305th 
Infantry. Two officers and two enlisted 
men landed shortly after 0830 on White 
Beach 1, which they proceeded to recon- 
noiter. The second group, the first organ- 
ized element of the division to come ashore, 
was the reconnaissance party of the 305th 
Field Artillery Battalion. At 0830 this 
group was boated in two DUKW's at the 
line of departure, where it remained for 
several hours. Finally, the vehicles were or- 
dered ashore, and by midafternoon the re- 
connaissance party had established itself on 
White Beach 2. 19 

16 J St Provisional Marine Brig Opn Plan 7-44, 
1 1 Jul 44. 

17 1st Provisional Marine Brig War Diary, p. 
4; 22d Marines SAR, p. 3; Lodge, Guam, pp. 

1 s 1 st Provisional Marine Brig Opn Plan 7-44, 
1 1 Jul 44. 

19 305th Fid Arty Bn, Action Against the En- 
emy, Guam, 2 1 Jul 44, p. 1 ; 243d Engr Combat 
Bn Hist, p. 1, in 77th Inf Div Opns Rpt Guam, 
2 1 Jul-10 Aug 44. 



Orcling Landing Craft. The sd Battalion of the, 305th Infantry waited three 
and a half hours for the signal to go in to the reef. 

Meanwhile, the infantrymen of Colonel 
Adair's 2d Battalion were climbing down 
cargo nets from their transports into the 
bobbing landing craft that were to carry 
them to the edge of the reef. By 1030 all 
boats were in position near the line of de- 
parture waking for the signal to go in. For 
three and a half hours they circled impa- 
tiently. At last, at 1405, came the message 
to proceed to the beach and assemble in an 
area 300 yards inland from Gaan Point. 20 

Unfortunately, no amphibian tractors 
were on hand to transport Adair's men 
over the reef and on to the shore line, and 
of course their LCVP's were too deep- 
drafted to negotiate either the reef or the 
shallow waters inland of it. Over the sides 
of their boats the men climbed, and waded 
the rest of the way in water at least waist 

20 1st Provisional Marine Brig War Diary, Jul 
44, p. 4; 30f,th RCT AAR, 18 Jun-9 Aug 44, 
p. 1. 

deep. Some lucky few were able to pick 
up rides in Marine LVT's on the land- 
ward side of the reef, but most stumbled 
in over the rough coral bottom, cutting 
their shoes en route and occasionally fall- 
ing into deep pot holes. Luckily, no enemy 
fire impeded their progress, and except for 
the dousing they got and the exhaustion 
they suffered the troops of the 2d Battalion, 
305th, completed their ship-to-shore move- 
ment without injury. 21 

The remaining two battalions of the 
305th Infantry encountered even greater 
difficulties in getting ashore. At 1530 
Colonel Tanzola received the message (a 

21 S. Sgt. James M. Burns, Narrative of the 
Guam Campaign, Notes, a i Jul 44 (hereafter 
cited as Burns Notes), MS, OCMIL Sergeant 
Burns headed the historical team attached to the 
77th Infantry Division on Guam. See also, AFAS, 
Guam, pp. 33-34. (This account was prepared in 
large part from the notes and interviews made by 
Sergeant Burns.) 



full hour after it had been dispatched) to 
land the rest of his regiment immediately. 
With only enough landing craft available to 
embark one battalion, he ordered Lt, Col, 
James E. Landrum, Jr., to boat his ist 
Battalion and head for the reef. However, 
naval officers in charge of boat traffic had 
not been apprised of these orders and re- 
fused to permit the landing craft to pro- 
ceed beyond the line of departure. Finally, 
at 1730, General Shepherd settled the mat- 
ter by ordering the regimental commander 
to land his combat team at once. Again 
no amphibian tractors were available and 
again the troops had to make their way 
through the water on foot. This time con- 
ditions were even more adverse. The tide 
was full, forcing some of the men to swim 
part of the way, and night was approach- 
ing. It was no easy task in view of the fact 
that the average infantryman was loaded 
down with a steel helmet and liner, gas 
mask, life belt, rifle, bayonet, grenade 
launcher, light pack, two bandoleers of am- 
munition, a bag full of rifle grenades hung 
from his neck, another pouch of hand 
grenades strapped to his thighs, a two-foot 
long pair of wire cutters tied to his pack, 
two canteens of water, a first aid pack, and 
a machete hanging from his cartridge 
belt. Nevertheless, the troops reached the 
beach safely after nightfall, although they 
were scattered and several hundred yards 
south of Gaan Point, where they were 
supposed to touch down. Colonel Landrum 
gathered the better part of his men to- 
gether and led them to their designated 
assembly areas inland. The rest had to 
spend the night on the beach, since after 
2130 Marine military police stopped all 
movement, 22 

The 3d Battalion, 305th, commanded by 
Lt. Col. Edward Chalgren, Jr., took even 
longer to get to shore. Before the landing 
craft assigned to the ist Battalion had re- 
turned to the transport area, the report 
came in that an enemy submarine had 
been spotted in the area, and all the ships 
of Transport Division 38 put to sea — in- 
cluding Alpine, aboard which Chalgren's 
unit was embarked. Not until 2 1 20 did the 
ship return, and it was 0330 of the 2 2d 
before the bulk of the battalion completed 
the complicated transfer from ship to land- 
ing craft to shore. To expedite the ship-to- 
shore movement of his battalion, Chalgren 
borrowed five LVT's from the Marines. 
One of these acted as control boat while 
the other four ferried the troops across the 
reef in driblets. Even so, some elements 
failed to reach the beach until 0600, and 
not for another hour was the unit fully 
reorganized. Meanwhile, Colonel Tanzola 
and his staff had debarked and reached the 
reef line, where they, too, found no trans- 
portation waiting them. The regimental 
commander came upon an abandoned rub- 
ber raft, commandeered it, and paddled his 
way to shore while the rest of his staff 
waded and swam in. 23 

Japanese Counterattack 

The American troops that dug in along 
their beachhead perimeters on the night of 
21-22 July were well aware that the odds 
favored an organized enemy counterattack, 
probably in the early morning hours. Jap- 

22 Burns Notes, 21 Jul 44; 305th RCT AAR, 18 
Jun-9 Aug 44, p. 1 ; 305th RCT Jnl Record Book, 

1 Jul-8 Sep 44, 21 Jul 44; ist Provisional Marine 
Brig Opus Jnl, 21 Jul 44, in ist Provisional Ma- 
rine Brig Opns and SAR; AFAS Guam, pp. 34- 
36; Ltr, Col Landrum to Gen A. G. Smith, 27 
Apr 55, Incl, OGMH. 

23 Burns Notes, 21 Jul 44; Lodge, Guam, pp. 



anese defensive doctrine prescribed such 
countermeasures and at both Saipan and 
Tinian the island commanders had rigidly 
adhered to doctrine. All along the line the 
troops were alerted to prepare for the ex- 
pected attack and took such measures as 
the terrain and their own dispositions per- 
mitted. When the counterthrust did ma- 
terialize, it was surprisingly feeble, at least 
in comparison with those that the enemy 
had launched against the invaders of Sai- 
pan and Tinian. 

In the zone of the 3d Marine Division, 
whose hold on the beachhead was the most 
precarious, no critical threat to the Marine 
lines occurred. Mortar and artillery fire 
falling on the beaches put a temporary end 
to unloading about 0230, and enemy pa- 
trols probed the front lines throughout the 
night, but that was all. 24 

Farther south, the Japanese undertook 
more serious measures. Against the Agat 
beaches, Colonel Suenaga of the 38th In- 
fantry Regiment himself led a force con- 
sisting of at least two infantry battalions 
supported by tanks. The brunt of the at- 
tack, which started about 0100, fell in 
the zone of the 4th Marines. In two suc- 
cessive waves, the enemy struck the Ma- 
rine perimeters around Hill 40 and suc- 
ceeded in penetrating as far as the pack 
howitzer positions on the beaches before 
being destroyed or beaten off. Another 
force, led by four tanks followed by guns 
mounted on trucks, came down Harmon 
Road in an apparent attempt to re-establish 
positions in the town of Agat. Fortun- 
ately, the marines had set up a roadblock, 
and a platoon of Sherman tanks with the 
help of a single bazooka quickly disposed 
of the Japanese armor, after which the 
enemy infantrymen retreated. About the 

24 Lodge, Guam, p. 59. 

same time, another group of Japanese hit 
the 1st Battalion's lines south of the road, 
and again succeeded in penetrating the de- 
fenders' lines as far to the rear as the ar- 
tillery positions. Fighting was so fierce in 
this area that one platoon of Company A 
was reduced to four men, 25 

It was here, too, that the soldiers of the 
305th Infantry got their first taste of close- 
in combat on Guam. One finger of the 
Japanese penetration reached the hastily 
established lines of Companies A and B, 
and in the fighting that ensued seven 
Americans were killed and ten wounded in 
exchange for about twenty Japanese. 26 

On their part, the Japanese lost at least 
268 men and 6 tanks during the counter- 
attack, 27 Among those killed was Colonel 
Suenaga who, after being wounded in the 
thigh by mortar fire, continued to lead the 
attack until he received a fatal bullet in the 
chest. Suenaga's death deprived the Japa- 
nese defenders on Orote Peninsula of their 
commander. Command of the sector passed 
to Comdr. Asaichi Tamai, IJN, of the 263d 
Air Unit, who took over not only the Navy 
forces stationed on Orote but also the sd 
Battalion, 38th Infantry, guarding the base 
of the peninsula. 38 

Consolidating the Southern Beachhead 
22-24 July 

General Shepherd's plan for 22 July con- 
templated an expansion of the beachhead 
in three directions — north, east, and south. 

™ Ibid., pp. 54-55' 

29 Burns Notes, 2i Jul 44. 

27 III Phib Corps Rpt Guam, Incl D, G-2 Rpt 
a. Major Lodge estimates the number of Japanese 
killed to be closer to 600. Lodge, Guam, p. 55. 

28 Ltr, Takeda to McQueen, 20 Feb 52. See 
also Takeda Rpt, 4 Oct 44; Japanese Studies in 
World War II, 55. 



On the right (south), the 4th Marines was 
ordered to seize Mount Alifan and then 
push south along the ridge line toward 
Mount Taene. The 22d Marines was to 
head in a northeasterly direction and cap- 
ture an intermediate objective line that ran 
about 4,000 yards east of Pelagi Rock, The 
305th Infantry (less 2d Battalion) was or- 
dered to pass through the left battalion of 
the 4th Marines and attack toward Maanot 
Pass, a cut through the mountains over 
which Harmon Road ran from Agat to the 
east. The 2d Battalion, 305th, and 2d Bat- 
talion, 4th Marines, were to be in brigade 

At the outset of the attack the 4th Ma- 
rines ran into heavy enfilade fire from the 
reverse slopes of Alifan's foothills. Once 
this was neutralized and the marines could 
approach the mountain itself, the chief de- 
terrent to rapid progress was the terrain. 
Approaches to the crest were snarled with 
pandanus roots and led to vertical cliffs 
covered with thick scrub growth. Enemy 
resistance, however, was only nominal, and 
one patrol was able to reach Alifan's sum- 
mit, where it discovered no evidence of 
Japanese activity. The patrol then returned 
to the foot of the cliffs and joined the rest 
of the regiment. 30 

In the brigade center the 305th Infan- 
try was slow in jumping off because of the 
many delays incurred in getting ashore the 
night before. Neither the 1st nor the 3d 
Battalion had been able to get fully organ- 
ized until about 0700; the regimental 
command post was not in a position to 
direct the attack until an hour later; the 
305th Field Artillery was even longer in 
getting set to go into action. By 1000, how- 

ever, all was in order and the regiment, 
with the 1 st Battalion on the right, the 
3d Battalion on the left, passed through 
the left flank of the 4th Marines and 
launched its attack to capture the high 
ground over which Harmon Road passed 
to the east. 31 

During the morning the 305th encoun- 
tered little resistance, and by 1252 the 
3d Battalion was already on the line that 
had been set as the day's objective. 32 That 
afternoon Chalgren's men pushed ahead 
still farther but before night were pulled 
back to their 1 300 positions in order to tie 
in with the 1st Battalion, which had made 
slower progress because of the dense un- 
derbrush that covered the hills over which 
it had had to advance. When the regiment 
dug in, its two forward battalions had firm 
contact, but a deep gully prevented phys- 
ical contact with the 2 2d Marines on the 
left. 33 

On the brigade left, the 2 2d Marines 
pushed off at 0845, ahead of the other two 
regiments. Until late afternoon, it encoun- 
tered only moderate resistance. The hill 
northeast of Agat that had held up the 
1st Battalion the day before was easily 
mounted as tanks cleared the way to the 
next high ground. When the marines on 
the left reached the Ayuja River they were 
temporarily halted when it was discovered 
that the Agat road bridge across the river 
had been destroyed. The men were reluc- 
tant to ford the river without their support- 
ing tanks, and the tanks could not cross be- 
cause the river banks were too steep. In 

20 1st Provisional Marine Brig Opns and SAR, 
War Diary, p. 5; 305th RCT AAR Guam, p. 2. 
30 Lodge, Guam, p. 66. 

31 305th RCT AAR, p. 2; 305th RCT Jnl Rec- 
ord Book 1 Jul-8 Sep 44, 22 Jul, Msgs 65, 66. 

32 305th RCT Jnl Record Book, 22 Jul, Msgs 
7i(c), 75- 

33 305th RCT AAR, p. 2; Burns Notes, 3/305, 
22 Jul 44. 



4TH Marines Moves Inland Toward Mt. Alifan 

the absence of engineers, amphibian tanks 
were called up to replace the tanks until 
provision could be made to get the latter 

This substitution of armored amphibians 
for tanks was contrary to prevailing Ma- 
rine Corps doctrine. Experience at Saipan 
had shown that amphibian tractors, ar- 
mored or otherwise, were extremely vulner- 
able to enemy fire when out of the water. 34 
Hence, that afternoon General Shepherd 
sent clear orders to all the units under his 
command that the amphibians should not 
be used inland except where absolutely 
necessary. 35 

In any event, the troops made their way 
across Ayuja River, some of them reaching 
a point beyond the Old Agat Road before 

34 See above j Ch. V] 

8B 1 st Provisional Marine Brig Opns and SAR, 
Jnl, 22 Jul 44, Msg 62. 

they were ordered back to the road itself 
to tic in for the night with the rest of the 
22d Regiment. Later, the 22d Marine Tank 
Company, after twenty-four hours of labor, 
constructed a causeway across the river to 
enable its vehicles to cross over. 

By nightfall the brigade, with the 305th 
Infantry attached, was drawn up along a 
line extending from the shore near Pelagi 
Rock, along the high ground northeast of 
Harmon Road, through Mount Alifan, 
southwest along the ridge line toward 
Mount Taene, then west to the sea at 
Magpo Point. The 4th Marines held this 
line from its coastal anchor at Magpo 
Point to Mount Alifan; the 305th Infan- 
try held the area along the foothills north- 
east of Mount Alifan to a point about 
1,200 yards from the Old Agat Road; and 
the 2 2d Marines held the area north and 
northeast of Agat. Approximately 3,000 



Assembly Area of 305TH Infantry on 22 July 

yards of the Force Beachhead Line in 
the brigade's zone was now in American 

It was clear by now that nothing was 
likely to impede the complete and rapid oc- 
cupation of the southern beachhead, and, 
accordingly, General Geiger announced the 
second objective for the brigade — the cap- 
ture of Orote Peninsula. Once the Force 
Beachhead Line had been fully occupied, 
the brigade was to be relieved by two regi- 
ments of the 77th Infantry Division, which 
would take over defense of the line from 
Inalas to Magpo Point. This task was to 
fall to the 305th Infantry, already ashore, 
and to Col. Aubrey D. Smith's 306th Reg- 
imental Combat Team, which was still 
afloat. The 305th would be released from 
brigade control, and both regiments were 

to be under direct command of General 
Bruce. The 307th was ordered to remain 
afloat in corps reserve and be prepared to 
land anywhere on order. 3 " 

The night of 22-23 J u ty m t^ e brigade 
zone was one of relative quiet, disturbed 
only occasionally by Japanese patrols, 
whose efforts to infiltrate through the lines 
were uniformly unsuccessful. The next 
morning, and throughout the 23d, the 4th 
Marines on the right (south) made no 
further effort to advance, but consolidated 
its positions and prepared to be relieved by 
the 306th Infantry. That these men were 
able to go about their preparations undis- 
turbed was partly due to the fact that 

3,,, 77th Inf Div G-3 Jnl, 22 Jul 44, Msg 1830, 
CG III Phib Corps to CG 77th Inf Div and CG 
1 st Provisional Marine Brig, 



naval guns and aircraft sealed off most of 
the possible routes of enemy counterattack. 
Early in the morning of the 23d observers 
had spotted a large Japanese troop move- 
ment headed from Mount Lamlam toward 
Facpi Point. The call went out immedi- 
ately for naval and air assistance, and in 
response the cruiser Honolulu, aided by 
carrier planes, dispersed the Japanese. 37 

For the other two regiments, the push 
toward the Force Beachhead Line was con- 
siderably facilitated by the landing of field 
and antiaircraft artillery. By the morning 
of 22 July, Brig. Gen. Pedro A. del Valle, 
USMC, had two thirds of the 155-mm. 
guns and howitzers of III Amphibious 
Corps Artillery ashore, and the 9th Marine 
Defense Battalion had succeeded in plac- 
ing two dozen 40-mm. and 20-mm. guns 
and a dozen .50-caliber machine guns 
along the beach between Agat and Bangi 
Point.* 8 Late the same afternoon, Brig. 
Gen. Isaac Spalding, Commanding Gen- 
eral, 77 th Division Artillery, set up his 
command post inland of Gaan Point. By 
then the remainder of the 305th Field Ar- 
tillery Battalion was ashore, as were two 
other artillery battalions, the 304th and 
902d, The 305th was ordered to continue 
to support the 305th Infantry; the 304th 
Field Artillery Battalion was assigned to 
close support of the 306th Infantry when 
it should get ashore; and the 902d was 
given the mission of general support for 
the division. The 304th and 902d Battal- 
ions were also ordered to be prepared to 
support the brigade in its forthcoming at- 
tack on Orote Peninsula. 39 

37 Lodge, Cuam, p. 66. 

S8 Ibid. 

3n 77th Inf Div Opns Rpt Guam, p. 3; 77th 
Inf Div Arty, Daily Summary of Events, S3 Jul 
44. P- 4- 

Meanwhile, on the 23d the 305th In- 
fantry had made rapid progress, hampered 
only by occasional enemy patrols. By 1 1 30 
the 3d Battalion, on the left, had reached 
the O— 2 line; the 1st was pushing from the 
Harmon Road along the ridge northeast of 
the Maanot Pass toward the O-2 line on 
the right of the 3d. The 2d Battalion 
moved to Maanot Pass to block the path 
of possible Japanese approach from the 
cast along the Harmon Road. By early af- 
ternoon, as the 305th Infantry was digging 
in for the night in a deep salient along the 
O-2 line and the Force Beachhead Line, 
all of the high ground overlooking Orote 
Peninsula from the southeast was in Amer- 
ican hands. 40 

To the left, the 2 2d Marines moved more 
slowly in its northward thrust toward the 
base of Orote Peninsula. Little resistance 
was encountered on 23 July until midafter- 
noon, but then, as the regiment attempted 
to swing across the narrow neck leading 
out onto the peninsula, the Japanese sud- 
denly came to life. The southern ap- 
proaches to the peninsula were guarded by 
marshy rice paddies interspersed with 
small hills that the Japanese had or- 
ganized into a system of mutually sup- 
porting strongpoints. Moreover, the zone of 
the 22d Marines approach was exposed to 
artillery and mortar fire from both Orote 
Peninsula on the west and Mount Tenjo 
to the northeast. The Japanese had not 
failed to make full use of these command- 
ing positions. In spite of heavy concentra- 
tion laid down by corps and division 
artillery, assisted by the big guns of the 
battleship Pennsylvania, the marines could 
do nothing to dislodge the enemy. After a 

40 305th RCT AAR, p. 2; 305th RCT Jnl, 
22 Jul 44, Msgs 93, 96, 102, 104b, 109. 



Smouldering Japanese Tanks knocked out on Agat—Sumay Road by US. 
medium tanks. 

wasted, exasperating afternoon sloshing 
through the rice paddies, the regiment fin- 
ally drew back a full 400 yards from its 
most advanced positions and dug in just 
south of the Old Agat Road. 41 

General Shepherd's plan for action on 
the 24th contemplated a two-pronged 
movement, the object of which was to en- 
velope the rice paddy area and establish a 
firm line across the narrow neck of Orote 
Peninsula. All three battalions of the 2 2d 
Marines were to be committed to the at- 
tack. On the right, the 2d Battalion was to 
jump off from its night positions south of 
the Old Agat Road, skirt to the right of 
the marshland, and push north to the town 
of Atantano, just inland of Apra Harbor. 

11 1 st. Provisional Marine Brig Opns and SAR, 
War Diary, 23 Jul 44, p. 7; Ibid., Jnl, 23 Jul 
44, Msgs 6a, 63b; Lodge, Guam, p. 67. 

On the left, the other two battalions were 
to drive north up the coast along the Agat- 
Sumay Road in columns of battalions. 
Once the rear battalion had reached a 
point on the Agat-Sumay Road about 600 
yards from its line of departure, it was to 
fan out to the right and then head north 
toward the south shore of Apra Harbor. 
The lead battalion would continue in a 
northeasterly direction along the Agat- 
Sumay Road. If this plan succeeded, night- 
fall would find the regiment drawn up 
across the neck of Orote Peninsula from 
the seacoast to Apra Harbor, with all three 
battalions abreast. Enemy defenses in the 
rice paddy area would be enveloped and 
presumably destroyed or neutralized. To 
relieve the 2 2d Marines of the necessity of 
lengthening its already thinning lines, the 
305th Infantry was ordered to shift left 



Maj. Gen. Andrew D. Bruce discuss- 
ing operational plans with Col, Doug* 
las McNair. 

as far as Old Agat Road, which was to be 
the new regimental boundary line. 42 

After delaying an hour to permit a 
longer preliminary naval bombardment, 
the regiment jumped off as ordered at 
iooo, 24 July. The battalion on the right, 
was counterattacked almost immediately 
and had to fall back to its previous night's 
positions south of the Old Agat Road. By 
early afternoon, however, it regained the 
initiative and was quickly able to reach its 

12 1st Provisional Marine Brig Opns and SAR, 
War Diary, p. 8 ; 1st Provisional Marine Brig Opn 
Order 15. (See Lodge, Guam.:, Maps 14, 15,) 

objective at Atantano, where it established 
visual contact with the 305th Infantry on 
its right. The other two battalions found 
the going slower as they felt their way up 
the Agat-Sumay Road. The road was 
thickly mined with aerial bombs and sub- 
ject to heavy enemy artillery fire from 
Pelagi Rock and Orote Peninsula. Five 
Japanese tanks appeared and contested the 
passage but were quickly disposed of by 
the heavier Shermans that were spearhead- 
ing the drive. As planned, on reaching the 
point on the road where the ground to the 
right permitted a decent foothold, the rear 
battalion fanned out in that direction. 
Without too much difficulty it overran the 
strongpoints that had held up the regiment 
on the 23d, but as evening approached and 
the battalion was compelled to dig in, it 
was still some 400 yards short of its objec- 
tive on the south shore of Apra Harbor. 
Meanwhile, on the left, the lead battalion 
had succeeded in reaching its objective on 
the west coast of the narrow neck of 
Orote, 43 

Thus by the end of 24 July, Gen- 
eral Shepherd's major objectives had been 
achieved. The Force Beachhead Line had 
been captured and was now manned 
mostly by units of the 77th Division. Orote 
was scaled off and the time was at hand 
to launch the drive to capture this vital 
peninsula that pointed fingerlike into the 
ocean between the two landing beaches and 
would constitute a danger to each so long 
as it remained in enemy hands. 

Landing the Reserves 

General Geiger's plan to commit the ma- 
jor portion of the 1st Provisional Marine 

48 1st Provisional Marine Brig Opns and SAR, 
War Diary, p. 8 ; Lodge, Guam, pp. 68-69. 



Brigade to the assault on Orote required 
that the 77th Division take over the duty 
of manning the Force Beachhead Line un- 
til the peninsula could be secured. Accord- 
ingly, he ordered the 305th Infantry to 
occupy the northern part of the line and 
the 306th to land and relieve the 4th Ma- 
rines on the southern part. 

Late in the afternoon of 22 July, Colonel 
Smith, commanding officer of the 306th, 
had come ashore with Col. Douglas C. Mc- 
Nair, 77th Division chief of staff, to recon- 
noiter the landing beaches and assembly 
areas. Next day the whole regiment made 
the damp journey from ship to shore. As in 
the case of the Army troops that had pre- 
ceded them, the men had to wade in from 
the reef line carrying all their equipment. 
The 77 th Division as a reserve unit had 
been assigned no LVT's and its DUKW's 
had to be reserved for cargo and light ar- 
tillery. Division headquarters had tried to 
laid the troops at low tide, but overloaded 
communications and other difficulties in 
procuring landing craft delayed the ship- 
to-shore movement until the tide was 
flooding. Vehicles had to be dragged by 
bulldozer from the reef to the beach, and 
most of them drowned out. Almost all of 
the radio sets, even those waterproofed, 
were ruined, and one tank disappeared al- 
together in a large shell hole. 

As a result of the delays incident to this 
slow procedure, only the 3d Battalion, 
commanded by Lt. Col. Gordon T. Kim- 
brell, was ashore in time to make its way 
into the line that afternoon. This battal- 
ion, with Companies A and C attached, 
relieved the 3d Battalion, 4th Marines, in 
its positions between Mount Alifan and 
Mount Taene where, after repulsing a 
small enemy attack, the Army battalion es- 
tablished its perimeters for the night. The 

other two battalions of the 306th stayed in 
their assembly areas. 44 

On the morning of 24 July Lt. Col. 
Charles F. Greene's 2d Battalion, 306th In- 
fantry, completed the relief of the 4th Ma- 
rines on the northern part of its line and 
established contact with the 305th Infan- 
try on the left. Throughout the day 
Greene's men cleaned out Japanese caves 
and dugouts in their area. No organized 
enemy appeared to contest their posi- 
tions. 45 

General Bruce therefore requested per- 
mission to land two battalions of the 307th 
Infantry so that he could push the Force 
Beachhead Line still farther east and as 
far south as Facpi Point. General Geiger 
denied the request because he considered 
a further expansion unnecessary and also 
because he was loath at this time to part 
with all but one battalion of his corps re- 
serve. 46 Nevertheless, he conceded that the 
307th should at least be landed in the area 
of the southern beachhead, although it 
was retained under the III Amphibious 
Corps command. Starting on the morning 
of 24 July, Col. Stephen S. Hamilton's 
regiment commenced the uncomfortable 
trek from its transports to the beach — 
made even more arduous this time by an 
untimely storm at sea. By midafternoon, 
General Bruce himself was ashore at his 
command post and ready to take over di- 
rection of his two regiments on the line. 47 

44 306th RCT Opn Rpt Guam, p. q; 306th 
RCT Jnl, 23 Jul-10 Aug 44, 23 Jul 44, Msgs a, 
4, 6; AFAS, Guam, p. 39; Ltr, Lt Gen A. D. 
Bruce, USA (Ret.), to Gen A. C. Smith, n Feb 
55, Incl 1, OCMH. 

4 r > 306th RCT Jnl, 24 Jul 44, Msgs 8, 1 -J, 14, 
15, 16; 306th RCT Opn Rpt Guam, p. a. 

1(i III Phib Corps Rpt Guam, Opns Rpt, p. 3. 

47 77th Inf Div Opns Rpt Guam, p. 5; 307th 
RCT Jnl, 24 Jul 44, Msg, 1030. 



Thus by the end of 24 July, after four 
days of fighting, the 1st Provisional Marine 
Brigade, with an assist from the 77 th In- 
fantry Division, had captured almost all 
the ground between the landing beaches 
and the Force Beachhead Line and had 
cut off from retreat and reinforcement the 
Japanese garrison on Orote Peninsula. The 
305th and 306th Regimental Combat 
Teams had taken over all of the line except 
that section across the neck of Orote Pen- 
insula, where the marines were poised to 
launch the next major assault. All of del 
Vallc's III Amphibious Corps Artillery was 
ashore, and most of the 77th Division Ar- 
tillery was in position to support the ad- 
vance. Beach defenses from Agat to Gaan 
Point were manned by the gth Marine De- 
fense Battalion, and from Gaan Point south 
to Bangi Point by the 7th Antiaircraft 
Artillery (Automatic Weapons) Battalion 
of the 77 th Division. Firmly in American 
hands was a large irregular slice of Guam 
running from Magpo Point northeast to 
Mount Tacnc, north to Mount Alifan, and 
again northeast to Inalas, where it bent 
back northwestward to Atantano and then 
southwest across the neck of Orote Pen- 
insula. Casualties up to the night of 
24 July were 188 killed in action, 728 
wounded, and 87 missing in the brigade; 
24 killed in action, 63 wounded, and 1 
missing in the Army division. 48 This was 
a substantial price to pay, but the figures 
were not out of proportion when compared 
to those for most Central Pacific cam- 
paigns. Nor were the losses nearly so great 
as those being suffered at the same time 
by the 3d Marine Division in its bitter fight 
for the northern beachhead. 

Consolidating the Northern Beachhead 
22-24 J u h 

Speaking of the terrain that faced the 
3d and 21st Regiments of the 3d Ma- 
rine Division just inland of their landing 
beaches, Lt. Gen. Alexander A. Vandegrift, 
commandant of the Marine Corps, later re- 
marked that it was "some of the most 
rugged country I have ever seen." 4M Com- 
ing as it did from a veteran of Guadal- 
canal, this description conveys some idea of 
the nature of the problem facing the ma- 
rines immediately upon their landing in 
this zone, Bundschu Ridge and Chonito 
Cliff were little more than a hundred feet 
in height, but rose precipitately from the 
narrow coastal plain beneath." Dense 
and tangled underbrush covered the ap- 
proaches to the ridge, and the whole area 
was striated with deep ravines and gul- 
leys that made contact between the attack- 
ing units extremely difficult to maintain 
and at the same time o ffered ideal r outes 
for Japanese infiltation. ^See Map VI.)\ 

On 22 July, while the 2ist Marines re- 
organized, the 3d Marines, on the left, as- 
saulted Bundschu Ridge behind a heavy 
concentration of artillery and heavy weap- 
ons, supplemented by the 20-mm. and 40- 
mm. guns of the 14th Marine Defense 
Battalion. A few men reached the top, but 
only to return to their parent unit before 
the day's close. By this time Colonel Hall's 
3d Marines was a badly battered regiment. 
In two days fighting he had lost 615 men, 
and some of his companies were down to 
30 or 40 men. 51 On the next day he de- 
cided to pit his 1 st and 2d Battalions, deci- 

48 Lodge, Guam, p. 70; 77th Inf Div G-i Jnl, 
24 Jul 44. 

i<J 3d Marine Div SAR, p. 4. 

50 See USAFICPA, Map, Island of Guam, 
Apra Harbor, Rev., Apr. 44. 

51 3d Marine Div SAR, p. a. 



mated as they were, against the ridge. To 
General Turnage he sent the message, "I 
am going to try and advance up that mess 
in front of me. What I really need is a 
battalion whereas I have only 160 men to 
use on that 500 yard slope." n2 To his great 
surprise, when his men finally hacked their 
way to the top on the morning of the 23d, 
most of the Japanese had evacuated. 

Still it was impossible to push on im- 
mediately. The cliffs were studded with 
machine guns that had to be mopped up, 
and contact had yet to be gained with the 
2 1 st Marines on the right. That regiment 
had made little progress during the day in 
the face of heavy opposition from enemy 
pillboxes. Colonel Hall asked for reinforce- 
ments from the 77th Division, whose 307th 
Regiment was still afloat, but General 
Gciger was unwilling to release his corps 
reserve.'"' 3 

On the morning of 24 July, two battal- 
ions of the 2 1st Marines attacked up the 
ravine that lay between the 21st and the 
3d Marines to the left. From the cliffs on 
either side came murderous machine gun 
fire. On top of this, American naval pilots, 
called in to make an air strike, were forced 
to drop their bombs so close to the attack- 
ing line that seventeen marines were killed 
or wounded. 54 That afternoon, two com- 
panies of the 3d Marines coming up from 
the left were able at last to establish con- 
tact between the two regiments. 

At the close of 24 July the left and cen- 
ter regiments of the 3d Marine Division 
had, by dint of great effort and many cas- 
ualties, succeeded at last in establishing a 
foothold on the first ridge line inland from 
the sea. Yet, though this victory was by no 

52 Quoted in Lodge, Guam, p. 62. 

ss HI Phib Corps Rpt Guam, Opns Rpt, p. 2. 

51 a 1 st Marines SAR, p. 3. 

means a Pyrrhic one, neither could it be 
considered decisive. Ahead still lay a scries 
of formidable ridges, running roughly par- 
allel to the sea, each one higher than the 
one before, and all culminating in the mid- 
island mountain mass, which in this area 
was dominated by Mount Chachao, Mount 
Alutom, and Mount Tenjo. 

Meanwhile, on the division right, the 
picture was not so discouraging. On 22 
July the 9th Marines moved southwest 
along the ocean shore against light resis- 
tance and occupied Piti Navy Yard. Late 
the same afternoon a minor amphibious 
assault of battalion size was sent against 
Cabras Island. The landing was unopposed 
and uneventful except that one LVT hit 
a land mine. The next day the occupation 
of the island was completed. On the 24th 
the regiment sent out a patrol south along 
the Piti-Sumay Road in an attempt to es- 
tablish contact with the brigade. Covered 
by amphibian tanks as they moved along 
the coastal road, the men eventually came 
under heavy machine gun fire from the in- 
land cliffs, and this, coupled with the fact 
that they were approaching the edge of the 
pattern of American fire being directed 
against Orote, persuaded them to return. 
Contact between the division and the bri- 
gade would have to await another day. 55 

Initial Supply Over the Beaches 

If the 3d Marine Division suffered more 
severely than the 1st Provisional Marine 
Brigade from enemy fire and nightmarish 
terrain, it did enjoy one advantage over 
the troops to the south — the movement of 
supplies from ship to shore during the early 
days of the assault was more expeditious. 

S5 Lodge, Guam, pp. 62-63. 



Pontoon Barge With Crane z.v a fueling station. Alligator is refueling from a 
barge inside the reef. 

Off the Asan beaches, the lagoon be- 
tween the shore line and the reef's edge was 
for the most part dry at low tide and 
trucks were able to drive to the edge and 
take on supplies directly from cargo boats 
and lighters. Off the Agat beaches, how- 
ever, the abutting reef and the lagoon bot- 
tom were underwater at all times, and 
consequently all cargo had to be restaged 
to amphibious vehicles several hundred 
yards offshore. To add to the difficulty, the 
fringe of the reef was not pronounced, and 
therefore it was impossible to position 
cranes on the coral outcroppings as was 
done at Asan. As a substitute measure, the 
cranes at Agat were mounted on pontoon 
barges moored along the line where the 
shallow water began. Using this system, it 
was possible for the cargo boats to load 
LVT's at the rate of 25 tons an hour and 

DUKW's at the rate of 40 tons an hour. 56 
Another device employed to expedite un- 
loading over the southern beaches was the 
construction of improvised pontoons made 
by lashing together two life rafts covered 
by a platform. Each attack transport of the 
Southern Transport Group was ordered to 
provide one of these "barges," each of 
which had a capacity slightly in excess of 
one ton. LCVP's could lower their ramps 
over the floating platforms and transfer